TW200529673A - Image encoding apparatus, image encoding method, image encoding program, image decoding apparatus, image decoding method, and image decoding program - Google Patents

Abstract

The image encoding apparatus of one embodiment of the present invention comprises a coding mode determination unit, a prediction image generation unit, a storage unit, and an encoding unit. The coding mode determination unit determines a coding mode relating to which of the first image prediction processing or second image prediction processing is used for generating prediction image of a partial area of input images. The prediction image generation unit extracts the prediction assist information by the first image prediction processing and generates a prediction image based on the prediction assist information. The storage unit stores the reproduced image that is based on the prediction image. The encoding unit generates a bit stream comprising data obtained by encoding the coding mode information and prediction assist information.

Description

200529673 (1) IX. Description of the invention [Technical field to which the invention belongs] The present invention relates to an image encoding device, an image encoding method, an image encoding program, an image decoding device, an image decoding method, and an image decoding program. [Prior art] In recent years, due to the popularity of the Internet, it is becoming more and more widespread to send or receive image data via the Internet, or to store image data. Generally speaking, the encoding of portrait data uses a forward encoding method which can effectively reduce the capacity of portrait data. As an example of this forward encoding method, there is an international standard specification H.264 animation encoding method disclosed by ITU-T (for example, refer to ITU-T VCEG (Q.6 / 16), "H. 26L Test Model Long Term Number 8 (TML- 8) draftO "). In the H.264 Internet coding frame (I frame), the animation is compressed by intra frame coding using intra prediction. In the internal frame coding of Η · 264, the frame to be coded is divided into 16x16 image-sized microblocks, and coding processing is performed in each microblock. The microblock is further divided into blocks of 16x8 pixels or 8x8 pixels, and motion compensation prediction is performed on each divided block. This reduces the length of time in the animation. ^ [Summary of the Invention] However, for the above-mentioned animation encoding such as animation encoding or still image encoding, a more efficient encoding technology will be obtained. 200529673 (2) Therefore, the object of the present invention is to provide an image encoding device, an image encoding method, and an image encoding program that can efficiently encode an image, and to provide an image that can be restored by using bits generated by the image encoding device of the present invention. Image decoding device, image decoding method, and image decoding program. One aspect of the present invention relates to an image encoding device including: (a) a plurality of partial fields formed by dividing an input image of an encoding object into a specific size, and determining the first Which of the 1 picture prediction process or the 2 picture prediction process generates a coding mode for predicting a picture, and a means for determining coding mode information for limiting the coding mode, and (b) executes a part of the aforementioned plural areas from The aforementioned encoding mode information is limited to a part of a field in which a predicted image should be generated by the first image prediction process, and the prediction subsidy information for generating the predicted image of the partial field from the already-played image in the other part field is extracted and based on the prediction. 1st image prediction means of the aforementioned first image prediction processing for generating the predicted image by the supplementary information, (c) a memory method of memorizing a playback image based on the foregoing image, and (d) generation of data including information on the encoding mode and the prediction assistance information The encoding method of the bit stream. The image encoding method related to the other aspect of the present invention, wherein the determining means includes: (a) For each of a plurality of partial fields formed by dividing the input image of the encoding object into a specific size, determining the relevant Which of the first image prediction process or the second image prediction process of the predicted image generates a coding mode of the predicted image, and generates a decision step for limiting coding mode information of the coding mode, (b) the first ... portrait pre-6 -200529673 (3) In the above-mentioned plural parts of the measuring means, from the aforementioned coding mode information, a part of the area where the predicted image should be generated by the first image prediction process is extracted, and the already-played image from the other part areas is extracted (1) The first image prediction step of the above-mentioned first image prediction process of generating the predicted portrait based on the predicted assistance information and the first portrait prediction step of the first portrait prediction process based on the prediction assistance information, (C) the memory means memorizing the playback portrait based on the foregoing portrait The memorizing step and (d) encoding means generate information including encoded encoding mode and prediction subsidy Encoding step of bit stream of information. In addition, the image encoding program related to the other side of the present invention uses a computer to function as: (a) For each of the plural fields formed by dividing the input image of the encoding object into a specific size, determine the relevant information through the need to predict the subsidy Which of the first image prediction process generating the predicted image or the second image prediction process generates the encoding mode of the prediction image, and the determination means for generating encoding mode information for limiting the encoding mode, (b) a plurality of partial fields, From the aforementioned coding mode information, a part of the field in which the predicted image should be generated by the first image prediction process is limited, and the playback image that has been generated from other parts of the field is extracted to generate prediction assistance information for the predicted image in that part of the field and based on the Prediction assistance information: The first portrait prediction method of the first portrait prediction process for generating the predicted portrait, (c) a memory method for memorizing a playback image based on the aforementioned portrait, and (d) generation of information including the encoded encoding mode information and the aforementioned prediction assistance. A program that encodes the bit stream of information. In addition, the above-mentioned image encoding program and the image encoding program of the present invention described below will be provided in the form of a computer-readable recording medium, a computer data signal or a program product superimposed on a carrier wave, etc. 200529673 (4). The above second image prediction processing is to set a part of the pixels where no prediction signal is generated as a template, and the generated playback image as a reference area, and select the reference area, and the area most relevant to the template is the copy reference. In the field, pixels that do not generate a prediction signal in the aforementioned template are given a pixel 値 of the corresponding pixels in the reference field, thereby generating a process of predicting an image.画像 Another aspect of the present invention relates to an image decoding device, which includes: (a) a first image prediction that is used to generate a predicted image from the inclusion of a coding area and a plurality of partial areas formed by dividing an image to be decoded into a specific size; Encoding mode information of the image processing or the second image prediction processing, and a bit stream of prediction assistance information that generates a prediction image based on the first image prediction processing, decoding means for decoding the encoding mode information and the prediction assistance information, (b) a complex number In some areas, from the decoding mode information, the prediction image in a certain area in which the prediction image should be generated by the first image prediction process is limited, and from the already played image, the first image is predicted by using the prediction assistance information. In the first image prediction means generated by the processing, (c) a plurality of partial fields, from the decoding mode information, the prediction image in a part of the area in which the predicted image should be generated by the second image prediction processing is limited, and by the second image prediction (D) The second image prediction means generated by the image prediction process, and (d) memorizing the playback image based on the predicted image Memory means; (e) In the second image prediction process, a field in which pixels that do not generate a prediction signal are a part is set as a template, and a playback image stored in the memory means is set as a reference field, and the reference field is selected to be related to the template. The highest field is the field of copy reference 200529673 (5). On the pixels that do not generate prediction signals in the template, the pixels of the corresponding pixels in the reference field are copied to generate the predicted image. A picture decoding method related to another aspect of the present invention includes: (a) a first picture prediction process for encoding a picture to be decoded for each of a plurality of partial fields formed by dividing a picture to be decoded into a specific size; Or decoding mode information of the second picture prediction process, and a bit stream of prediction assistance information of the predicted picture generated by the first picture prediction process, a decoding step to decode the encoding mode information and the prediction assistance information, (b ) And the first image prediction means, in the part of the plurality of fields, from the decoding mode information, the prediction image in a part of the field in which the prediction image should be generated by the first image prediction processing is limited, and the prediction assistance information is used The first image prediction step generated by the first image prediction process and (c) the second image prediction means are limited in a plurality of fields from the aforementioned decoding mode information to restrict the number of prediction images that should be generated by the second image prediction process. The predicted image of a part of the field is the second image generated by the second image prediction process The measurement step, (d) and the memorizing means memorize the memorizing step based on playing the predicted picture; (e) In the second picture prediction process, the area in which pixels that do not generate a prediction signal are part of is set as a template and memorized in the memorizing means The playback image is set as the reference area, and the selected area is the most relevant area of the template. The reference area is the copied reference area. On the pixels in the template that do not generate prediction signals, the corresponding pixels in the copied reference area are given. The picture element 値 'thus produces a predictive portrait. The image decoding program related to the other aspect of the present invention is a computerized function of -9-200529673 (6) as follows: (a) For each part of the complex number formed by dividing the image of the decoding object into a specific size, the limitation is from the included encoding The decoding mode information of the first picture prediction process' or the second picture prediction process for generating a predicted picture, and the bit stream of the prediction assistance information of the predicted picture generated by the first picture prediction process are used to decode the encoding mode information and Decoding means of the prediction assistance information, (b) Among the plurality of areas described above, from the decoding mode information, the prediction image in a certain area in which the prediction image should be generated by the first image prediction process is limited, and the prediction assistance information is used. In the first image prediction means generated by the first image prediction process, (e) a plurality of partial fields, from the decoding mode information, the prediction image in a part of the field in which the predicted image should be generated by the second image prediction process is limited (D) a second image prediction means generated by the second image prediction process, (d) memory based on the predicted image playback Procedures for the memorization of portraits. In this program (e), in the second image prediction process, a field in which pixels that do not generate a prediction signal are a part is set as a template, and a playback image stored in a memory means is set as a reference field, and the reference field and the aforementioned template are selected. The most relevant field is the copy reference field. On the pixels in the template where the aforementioned prediction signal is not generated, the pixels of the corresponding pixels in the aforementioned reference region are applied to generate the predicted image and the computer is functionalized. In addition, the image decoding program described above and the image decoding program of the present invention described below can be provided in the form of a computer-readable recording medium, a computer data signal superimposed on a carrier wave, or a program product. According to the present invention and the second image prediction processing, the generated playback image is set as the reference area, and the pixels on which no prediction signal is generated in the model -10- 200529673 (7) are copied from the reference area selected Copy the corresponding pixels in the reference area. From the reference fields, select the field that is more relevant to the template as the copy reference field. For example, choose the highest related field, or the highest related field compared to a specific benchmark. Therefore, in the decoding end, for some areas that are limited to the areas in which the predicted picture should be generated by the second picture prediction process, the predicted picture can be actively generated without using the prediction assistance information from the encoding side. Therefore, high-efficiency coding can be achieved at the coding end. In the present invention related to image coding, the determining means is preferably to generate a plurality of broadcast images in some fields through a first image prediction process and generate them according to a specific scanning order, and then follow the order opposite to the specific scanning order. Select a part of the processing object, compared to the part of the processing object, the image of the part in the front of the scanning order, compared to the part of the processing object, in the part of the scanning order. Based on the comparison of the predicted image field that should be generated by the first image prediction process, the playback image of that part of the coding mode is set as the reference field, and the second image prediction process is used to generate the playback image of the part of the processing object. Based on the comparison, The playback image of the partial area of the processing target generated by the second image prediction processing and the playback image of the partial area of the processing target generated by the first image prediction processing determine the encoding mode of the partial area of the processing target. In this case, in the present invention related to image decoding, the prediction image is generated by the first image prediction process in accordance with a specific scanning order for a part of the field that is limited from the decoding mode information to the prediction image that should be generated by the first image pre-processing. The second image prediction method is to limit the area where the predicted image should be generated by the second image prediction process from the decoding mode information, after the playback image generated based on the predicted image is stored in the aforementioned -11-200529673 (8) billion means. In some areas, a predicted image is generated by a second image prediction process in accordance with a specific scanning order. According to the present invention, after the first portrait prediction process generates a playback image, the generated playback image is set as a reference area for the second image prediction process. Therefore, the playback image located at $ in the specific scanning order can also be used in the second image prediction process, so the redundant length in the spatial direction can be effectively reduced. In the present invention related to image encoding, the determining means respectively calculates a playback image of a partial area of the processing target generated by the second image prediction processing and a playback image of a partial area of the processing target generated by the first image prediction processing. The cost of the cost function derived from the coding skew or / and the amount of coding data, based on the cost, determines the aforementioned coding mode in some areas of the processing object. In the present invention related to image coding, in a plurality of partial fields, from the aforementioned coding mode information, the predicted image in a certain area in which a predicted image should be generated by the second image prediction processing is limited, and the second image is used to predict The second image prediction means generated by the processing; the second image prediction means sets the predicted image generated by the second image prediction processing as a playback image. At this time, in the present invention related to image decoding, the second image prediction means sets the predicted image generated by the second image prediction processing to play a day image. That is, the predicted image generated by the second image prediction method is still adopted as the broadcast image. Therefore, the playback image generated by the second image prediction process can also be used for prediction in some areas of the next process -12- 200529673 (9), so the redundant length can be reduced. In addition, since the information related to the difference between the input image and the predicted image generated by the second image prediction process need not be included in the bit stream, more efficient encoding can be achieved. In the present invention, the input portrait of the encoding object may also be an animation frame. At this time, in the second image prediction processing, at least one of the playback image of the encoding target frame and the playback image of the frame prior to the encoding target frame is set as a reference area. In this case, in the present invention related to image decoding, the “picture to be decoded is an animated picture frame; the second picture prediction means is stored in the play picture of the aforementioned memory means, and the play picture of the aforementioned picture frame is decoded; and At least one of the playback images of the frame that has priority over the decoding target frame is set as the aforementioned reference area. According to the present invention, in the second portrait prediction processing, both the playback image of the encoding target frame and the playback image of the processed frame different from the encoding target frame are set as reference areas, so the time direction and The length in space. In the present invention related to image coding, the first image prediction process may be a motion compensation prediction process. In this case, the prediction assistance information includes a motion vector extracted by the first image prediction process. In this case, in the present invention related to picture decoding, the prediction assistance information also includes a motion vector used in the first picture prediction process. In the present invention, the first portrait prediction process may also be a process of generating a predicted image by playing an image in the prediction in the same space in a part of the domain of the processing object. That is, the first picture prediction process may be a prediction process used for encoding and decoding of a still picture, or an in-frame prediction process used for encoding and decoding of a movie -13- 200529673 (10). In the first image prediction process, the connected partial fields connected from the encoding mode to the partial fields of the processing target are limited to the case where the partial image of the predicted image should be generated by the second image prediction processing, based on the partial fields that are not related to the processing target. Played portraits of the connected non-connected partial areas produce predicted images of the processed partial area. According to the present invention, a part of the processing target area is a field where a predicted image should be generated by the first image prediction process, and a part of the field connected to the processing object is a field where a predicted image should be generated by the second image prediction process. At that time, it is also possible to use the broadcast image data of non-connected partial areas that are not connected to the partial area of the processing target to generate a predicted portrait. Thereby, not only the playback portraits of the connected partial areas connected to the processing target partial areas, but also the playback portraits of the non-connected partial areas not connected to the processing target partial areas can be used to generate the predicted portraits. Therefore, the reference range when generating predictive image data will be expanded, and the redundant length of space will be reduced, so the encoding efficiency can be further improved. In addition, in the present invention related to portrait coding, in the first portrait prediction process, a prediction mode for predicting a portrait in a part of a field to be processed is determined from a plurality of prediction modes related to a plurality of different prediction rules. It also generates prediction mode information that limits the prediction mode. The encoding means can include the data of the encoded prediction mode information in the bit stream. Here, in the present invention related to portrait decoding, in the case where a bitstream is used to generate a predicted portrait by the first portrait prediction process, the complex prediction mode related to prediction rules with different encoding complex numbers is limited to the first portrait prediction Data of prediction mode information of the processed prediction mode; decoding means decodes prediction mode -14-200529673 (11) type information from the bit stream. In the first portrait prediction process, a prediction portrait is generated based on the prediction mode information. In the present invention, in the first image prediction process, the part of the connection area connecting the processing target areas from the coding mode is limited to a part of the area where the Yuci image should be generated by the second image prediction processing, and the line is located in the prediction direction. It is best to generate the predicted portrait based on the pixel 値 of the pixels in the part of the area closest to the processing target in the playback portrait of the non-connected partial area in the direction of the prediction start. Thereby, from the non-connected blocks, the best playback image data corresponding to the pattern generated by each predicted image is selected. In the present invention related to image encoding, the prediction residual image generation means performs a calculation of the difference between the prediction image generated by the first image prediction method and the input image of the encoding target to generate a prediction residual image; the encoding means bases the encoding on the prediction The information of the signal generated by the residual image may be included in the bit stream. Here, in the present invention related to image decoding, the bit stream includes a signal that encodes a prediction residual image that is generated based on a prediction calculation of a partial image generated by the first image prediction process and a difference calculation between the partial image and the image in the partial field. The generated data; decoding means included in the bit stream data, decoding the signal from the data formed based on the signal generated based on the prediction of the residual image, and decoding the signal; the method of generating the image by addition is based on the decoding by the decoding method. Respond to the signal Predicted residual image and predicted image, to generate a playback image. In the present invention related to image coding, the predicted residual image generation means executes the predicted image generated by executing the first image prediction method, and the encoding target -15- 200529673 ( 12) Calculate the difference between the input images to generate the predicted residual image. The conversion means generates conversion information by performing the conversion process on the predicted residual image. The reverse conversion means generates the predicted residual image by performing the inverse conversion process on the conversion information. Image generation means predicts residual images and predictions by adding It is also possible to measure the image, generate the broadcast image, and the encoding means may include the data of the code conversion information in the bit stream. Here, in the present invention related to image decoding, the bit stream includes prediction residues generated by calculating the difference between the prediction image of the partial field generated by the first image prediction process and the portrait (input image) of the partial field. Data generated by performing conversion processing on the image. Decoding means decodes the conversion information from the bit stream. Inverse conversion means generates a residual prediction image by performing a reverse conversion process on the conversion information. Playing back the image generation means by addition The predicted residual image and the predicted image are returned to generate a playback image. [Embodiment] Hereinafter, the best implementation mode of the present invention will be described in detail with reference to the drawings. In addition, in each drawing, the same or equivalent part will be marked with the same symbol. [First Embodiment] First, a first embodiment of the present invention will be described. Fig. 1 is a block diagram showing a picture encoding device according to a first embodiment. The image encoding device 1 shown in FIG. 1 can be physically a computer provided with, for example, a CPU (Central Processing Unit), a memory device such as a memory, a display device such as a display, and a communication device. The image coding device 1 may be a mobile communication terminal such as a mobile phone or a DVD device. That is, the image encoding device 1 can be widely used as an information processing device. The image encoding device 1 is functionally provided with an image segmentation unit (image segmentation means) 1 1. An encoding mode determination unit (determination means) 1 2. A predicted image generation unit (first image prediction means) 1 3. A subtraction unit (predicted residual image generation) Means) 14. Conversion unit (conversion means) 1 5. Encoding unit (encoding means) 1 6. Anti-conversion unit (anti-conversion means) 1 7. Addition unit (playback image generation means) 1 8 and memory unit (memory means) 1 9. Next, each structural element shown in FIG. 1 is demonstrated. The image segmentation section 11 divides the input image input in the frame unit into blocks of a certain size (for example, 4 X 4 portraits), that is, some areas. In addition, the image segmentation unit U generates block position information for limiting a processing object block that is an object of encoding processing. As the block location information, for example, the blocks in the frame are in the order of raster scanning, and the block numbers of 0, 1, and 2 are numbered from large to small, or a map containing each block is included. The upper left of the box is the block coordinates shown by the coordinates of the datum. The coding mode decision unit 12 determines the coding mode of each block based on a specific coding mode decision rule, and generates coding mode information for limiting the coding mode. ... In this embodiment, in the encoding mode, there is a predictive encoding processing mode (P mode) for encoding the processing target area by using the input image of the processing target block and the predicted image corresponding to the image. 'Without adopting. The input image of the processing target block and the predicted image corresponding to the image, encoding-17- 200529673 (14) Complementary coding processing mode (C mode) of the portrait mode of the processing target block. That is, when the encoding mode is a predictive encoding processing mode, information related to the image of the encoding processing target block is output. On the other hand, when the coding mode is the complement coding processing mode, the image related information of the processing target block is not coded, nor is it output. When the encoding mode is the predictive encoding processing mode, the first portrait prediction processing for generating the prediction assistance information required by the prediction image on the decoding side generates the predicted image, and when the encoding mode is the complement coding processing mode, decoding is performed. 2nd image prediction processing (image complementation processing) without prediction assistance information to generate predicted images. The encoding mode decision rules are generated, for example, by the image complementation processing described later (refer to FIG. 6 and the second image prediction processing). Processing) to generate the playback image of the processing target block, and set the square of the error between the input image of the processing target block and the playback image. The encoding processing mode is also available. In addition, it is not necessary to compare the square of error 値 and critical 値, but also the absolute 値 and critical 値 of error. In addition, other coding mode determination rules may be, for example, that information of a block that is coded in a predictive coding processing mode and a block that is coded in a complement coding mode is preliminarily corresponded to and stored in the block position information. During processing, the coding mode corresponding to the processing target block is obtained based on the block position information. When the encoding mode is the predictive encoding processing mode, the predicted image generation unit 13 generates the predicted image in the frame used for generating the predicted image corresponding to the input image of the processing target block by the first image prediction processing. -18- (15) (15) 200529673 Generate patterns, that is, prediction modes, select patterns from the nine types of prediction portraits described below, and output prediction mode information used to select the prediction mode. That is, the prediction image generation pattern (prediction mode) in this frame, and the prediction assistance information required when the decoding side generates the prediction image. The predicted image generation unit 13 generates a pattern according to the determined predicted image, and uses a part of the playback image that is fully encoded, played, and stored in the memory 19 in the image of each block to generate an input image corresponding to the processing target block. Of predictive portraits. The details of the first image prediction process when a predicted image is generated will be described later. The subtraction unit 14 subtracts the predicted image of the processing target block from the input image of the processing target block by the pixel unit to generate a predicted residual image. The conversion unit 15 converts the predicted residual image using a specific conversion rule, and outputs a conversion coefficient (conversion information) obtained by the conversion. Specific conversion rules are, for example, 2-dimensional element DCT with 4 rows and 4 columns, and orthogonal conversion and quantization with 4 rows and 4 columns used by H.264. The specific conversion rule may be a conversion operation or quantization such as matching pursuit, vector quantization, and wave rate conversion. The encoding unit 16 encodes a conversion coefficient based on a specific rule average message amount. The encoding unit 16 encodes the encoding mode information and the predicted picture generation pattern (prediction mode) based on a specific rule to average the amount of information. The average information amount coding is, for example, arithmetic coding. The inverse conversion unit 17 inversely converts the conversion coefficient with a specific inverse conversion rule to generate a revertive prediction residual image. This specific anti-conversion rule corresponds to the anti-conversion rule of the specific conversion rule adopted by the conversion department 15 (19) 200529673. The adding unit 18 is to add a prediction image corresponding to the block to be processed and a predicted residual image corresponding to the predicted image to generate a playback image. When the adding unit 18 sets a specific range of the picture element 値, the addition unit 18 may perform clipping in order to limit the picture element 内 to a specific range. The memory unit 19 records the playback image generated by the addition unit 18 in billions of records (not shown). Next, the first picture prediction process when a predicted picture is generated will be described with reference to Figs. 2 and 3. In this embodiment, the first image prediction process is an in-frame prediction process, but in the first image prediction process, various prediction processes such as motion compensation prediction process can be applied. First, as shown in FIG. 2, the block adjacent to the upper left of the processing object block Y of 4x4 pixels is set to block X0, and the block adjacent to the upper side is set to block XI. The block adjacent to the top is set to block X2, and the block adjacent to the left is set to block X3. Further, a block adjacent to the upper side of the block X1 is referred to as X4, a block adjacent to the upper side of the block X2 is referred to as a block X5, and a block adjacent to a left side of the block X3 is referred to as a block X6. In addition, the bottommost playback pixel 値 of block XI is A, B, C, D from the left, and the bottommost playback pixel 値 of block X2 is E, F, G, H from the left. The playback pixels 最 in the rightmost column of block X3 are, from the top, I, J, K, and L. In addition, the playback pixel 値 in the lower right corner of block X0 is set to M. Furthermore, the processing of the pixel 値 of the predicted image of the image block Υ is set to a, b, c, d, e, f, g, h, i, j, k, 1 according to the scanning order of the light spots. , M, η, ο, ρ〇 Here, reference will be made to FIG. 3 for the 4x4 pixel block coding mode 9 -20- 200529673 (17)

The prediction modes A 0 to A 8 will be described. The prediction mode A0 shown in FIG. 3 (a) is a mode in which a pixel 値 adjacent to the upper side of the processing target block is extended straight below, thereby generating a predicted portrait. In this prediction mode A 0, a prediction portrait is generated based on the following expression. a = e = i = m = A b = f = j = n = B c = g = k = o = C The prediction mode A1 shown in Fig. 3 (b) is obtained by placing adjacent to the left side of the processing object The picture element 値 is a model that extends straight to produce a predicted picture. In this prediction mode A1, a prediction image is generated based on the following expression.

3. — b — c — d-I e = f = g = h = J i = j = k = 1 = K m = n = o = p :: = L The prediction mode shown in Figure 3 (c) A2 is a model that uses the average value of surrounding pixels to predict only the DC component of the processing target block. In this prediction mode A2, a prediction image is generated based on the following rules. First, when A ~ M are the playback pixels in all the frames, all the numbers of a ~ ρ are set to (A + B + C + D + I + J + K + L + 4) / 8. In contrast, in the case where A ~ D does not belong to the playback block in the frame, and I ~ L belongs to the playback block in the frame, all the numbers of a ~ p are set to (I + J + K + L + twenty four. In addition, I ~ L does not belong to the playback block in the frame, and A ~ D belongs to the playback block in the frame. All numbers of a ~ ρ are set to (A + B + C + D + 2) -21-200529673 (18) The sentiment of the block is adjacent to the straight line of processing, based on the / 4. In addition, none of A to D and I to L belong to the broadcasting conditions in the frame, and all numbers of a to p are set to 128. The prediction mode A3 shown in Fig. 3 (d) is a mode for generating a predicted portrait by extending the pixels on the upper side of the similar block and the diagonally above the right to the diagonally below and to the left. In this prediction mode A3, a prediction image is generated as follows. a = (A + 2B + C + 2) / 4 b two e = (B + 2C + D + 2) / 4 c = f = i = (C + 2D + E + 2) / 4 d = = m = (D + 2E + F + 2) / 4 h = k = n = (E + 2F + G + 2) / 4 l = o = (F + 2G + H + 2) / 4 p = (G + 3H + 2 ) / 4 In the straight line A4 adjacent to the diagonally below the processing pair, the prediction mode A4 shown in Fig. 3 (e) is based on extending the pixels on the left, diagonally above and above the similar block to the right. A model for generating predictive portraits. In this prediction mode, a prediction image is generated in the following formula. m = (J + 2K + L + 2) / 4 i = n = (I + 2J + K + 2) / 4 e = j = o = (M + 2I + J + 2) / 4 a = f = k = p = (A + 2M + I + 2) / 4 b = g = 1 = (M + 2A + B + 2) / 4 c = h = (A + 2B + C + 2) / 4 d = (B + 2C + D + 2) / 4 -22 · 200529673 (19) The prediction mode a5 shown in Fig. 3 (f) is by arranging the pixels adjacent to the left side, the upper left side and the upper side of the processing target block. A model that stretches straight on the lower right side to produce a predicted image. In this prediction mode A5, a prediction image is generated based on the following formula. b = k = c 2 1 = (M + A + 1) / 2 (A + B + 1) / 2 (B + C + 1) / 2 d = f = o = (C + D + 1) / 2 (M + 2A + B +2) / 4 g = P = (A + 2B + C + 2) / 4 h = i two m = (B + 2C + D + 2) / 4 (M + 21+ J + 2) / 4 (1+ 2J + K + 2) / 4 The prediction mode A6 shown in Fig. 3 (g) is to produce a predictive portrait by extending the pixels adjacent to the left side, the upper left side and the upper side of the processing target block in a straight line to the lower right side. mode. In this prediction mode A6, a prediction image is generated based on the following formula. a = g = (M + I + 1) / 2 b = h = c = (I + 2M + A + 2) / 4 (M + 2A + B + 2) / 4 d = (A + 2B + C + 2) / 4 e = k = f = 1 = i = o = (I + J + 1) / 2 (M + 21+ J + 2) / 4 (J + K + 1) / 2 -23- 200529673 (20) (I + 2J + K + 2)-/ 4 (K + L + 1) / 2 η- (J + 2K + L + 2) / 4 The prediction mode A7 shown in Fig. 3 (h) is based on the The pixels 値 on the upper side of the object block and the upper right obliquely extend in a straight line on the lower left side to generate a model of predicted portraits. In this prediction mode A7, a prediction portrait is generated based on the following formula. a = (A + B + 1) / 2 b = i = (B + C + 1) / 2 c = j- (C + D + 1) / 2 d = k = (D + E + 1) / 2 1 = (E + F + 1) / 2 e = (A + 2B + C + 2) / 4 f = m = (B + 2C + D + 2) / 4 g = n = (C + 2D + E + 2) / 4 h- 〇 = (D + 2E + F + 2) / 4 p = (E + 2F + G + 2) / 4 The prediction mode A8 shown in Fig. 3 (i) is by placing adjacent to the processing target block The picture element on the left is a pattern that extends straight on the upper right side to produce a predicted picture. In this prediction mode A8, a prediction image is generated based on the following formula. (I + J + 1) // 2 b = (I + 2 J + K + 2) / 4 c = e = (J + K + 1) / 2 -24- 200529673 (21)

d = f = (J + 2K + L + 2) / 4 g = i = (K + L + 1) / 2 (K + 3L + 2) / 4 k = 1 = m = n = o -p- L prediction Image generation unit: 1 3. In each of the above prediction modes, even if the playback pixel 値 used when generating the prediction image is 1, the prediction mode outside the frame is not selected. Here, in this embodiment, if the block encoding mode of any of A to M is the complement coding mode, no predictive image is generated, and no pixels are played at this time. Therefore, when predictive portraits are generated in other blocks, it is not possible to refer to the block to play pixels in the complementary coding mode. In this embodiment, in this case, the line of the prediction direction of the prediction mode (the direction of the arrow shown in FIG. 3) and the direction of the prediction start side (the direction of the start of the arrow shown in FIG. 3) exist. Among the playback pixels in the same picture frame, the playback pixels that are closest to the processing target block are used as A to M instead of the pixel to generate a predicted image. When the straight line of the prediction direction passes between two day pixels, the average pixel 値 of the two pixels is set to replace the pixel 代替. The above-mentioned alternative pixel 値 will be described with reference to FIGS. 4 and 5. First, FIG. 4 is a diagram showing a process of replacing pixels determined when the block X 1 adjacent to the upper side of the block γ is a complement coding mode and the pre-tank mode is A0. As shown in FIG. 4, the playback pixels 播放 N, 0, P, and Q at the bottom of the block X4 are selected as the substitute pixels 値 of the playback pixels 値 A, B, C, and D of the block XI. That is, for the playback pixels 値 A, B, C, and D, on the straight line of the prediction direction of the prediction mode-25- 200529673 (22) A0, in the playback pixels 位于 located in the same frame as the prediction start side, Select the playback pixel 接近 that is closest to the processing target block Y > ^, 0,? , (5. Therefore, when the predicted image of the processing target block Y shown in FIG. 4 is generated, N, 0, P, and Q will be used to replace A, B, C, and D in the above-mentioned prediction mode A0. In addition, The block X4 is outside the frame or the complement coding mode, and A, B, C, and D are set outside the frame. That is, the prediction mode will not select A0. Second, Figure 5 shows the processing for block Y. The block XI adjacent to the upper side is a complement coding mode, and the prediction mode is A7 instead of the determined pixel 値 map. As shown in FIG. 5, the two bottom pixels 値 N, 〇 and the leftmost two playback pixels 区块 P, Q of block X2 are used as alternatives to the playback pixels 値 A, B, C, and D of block X1. Furthermore, N is selected instead of A値, choose 0 as a substitute for B, choose P as a substitute for D, and use the average Ο of P and P as a substitute for C. Use the average Ο of P and P as a substitute for C, because of the prediction direction. A straight line passes between the two pixels 0 and P (located on the pixel in the lower left corner of block X5). That is, the prediction method of the playback pixels 値 A, B, C, and D and the prediction mode A7 Among the playback pixels 直线 on the straight line and located in the same frame of the prediction start side, select the playback pixels 値 N, 〇, (〇 + p) that are closest to the playback pixels 値 of the processing target block Y. ) / 2, Q. Therefore, when the predicted portrait of the processing target block γ shown in FIG. 5 is generated, N, 0, (〇 + P) / 2, Q will be used to replace the above-mentioned prediction mode A7. A, B, C, D. In addition, when the average value of two pixel values is calculated in binary, 'add 2 pixel values, and then add -26- 200529673 (23) After adding 1 to the result of the method, 1 Bits can also be shifted to the right. In this way, by deciding to replace pixel 値, the best playback pixel corresponding to each prediction mode can be selected from non-adjacent blocks. Second, referring to FIG. 6, and A description will be given of the image complementation processing (second image prediction processing) performed when the supplementary portrait is generated in the above-mentioned coding mode determination rule. In the image complementation processing of this embodiment, as shown in FIG. One pixel contained in the target block Y is set as the processing target pixel P. The processing target pixel P and The area in which the picture pixels (playing pixels) of the picture are played near the object pixel P is set as a template T (template). In addition, there are picture pixels (pictures) in the picture Y which have been subjected to the picture complementing process. In the case of a picture element, the picture element that has been subjected to the image complementation process may be included in the template T. In addition, the process target block Y and the surrounding area of the process target block Y are set as the target area R.

First, the processing target block Y is scanned based on a specific scanning rule, thereby selecting a processing target pixel P from a plurality of pixels included in the processing target block Y. Next, the template T is determined based on the selected processing pixel P. Next, among the target area P, among the areas having the same shape as the template T, a relevant area S having the greatest correlation with the removal of the pixel of the processing target pixel P from the template T is selected. Secondly, the pixel 値 of the playback pixel Q corresponding to the processing target pixel P in the related field S is set to 塡 complement pixel 値 of the processing target pixel P. The same processing as above is performed for each pixel included in the processing target block Y in the scanning order. As a result, a supplementary portrait corresponding to the processing target block Y is generated. Here, the correlation calculation method when selecting the relevant field S -27- 200529673 (24) is selected, for example, the method with the smallest squared difference between the corresponding pixels 値 is set to the maximum correlation method, or the corresponding images The absolute sum total of the difference between the primes is the smallest, and the method of setting the maximum correlation is also possible. Any other method that can measure the correlation is also applicable. In addition, in the case where the encoding target image is an animation, by using the decoded pixels of the decoded frame and the complementary image as the target area R, the complementary image can be more efficiently followed by referring to FIG. The operation of the image encoding device 1 will be described. This image encoding process is performed from a frame according to a specific scanning order (such as light spot scanning) in the unit of the read block. First, the image segmentation unit 11 divides the input image input in the frame unit into blocks of a specific size (for example, 4 X 4 pixels), and generates block position information for limiting each processing target block (step S1). ). Next, the encoding mode determination unit 12 determines whether the encoding mode for encoding the image of the processing target block based on the specific encoding mode decision rule is one of a predictive encoding processing mode (P mode) or a complement encoding mode (C mode). And output coding mode information for limiting the coding mode (step S2). This block encoding mode information is output to the image division section 11, the prediction image generation section 13 and the encoding section 16. Next, the predicted image generation unit 13 determines whether the encoding mode of the processing target block determined by the encoding mode determination unit 12 is a predictive encoding processing mode (step S3). 9 If it is judged as negative (step S3; NO), in order to perform the second image encoding of the processing target block -28- 200529673 (25) • processing, the process moves to step S11. On the other hand, in the judgment of step S3, it is determined that the encoding mode of the processing target block is a predictive encoding processing mode (step S3; YES), and the predictive image generation unit 13 determines a predictive mode, and follows the predictive mode determined according to this. A part of the playback image that has been encoded and stored in the memory 19 is used to generate a predicted image corresponding to the image of the processing target block (step S4). That is, the predicted image generation unit 13 performs the first image prediction process described above, and generates a preview image based on the connected image connected to the processing target block and the playback image of the non-connected block not connected to the processing target block. •• Take a picture. This predicted image is output to the subtracting section 14 and the adding section 18. Next, the subtracting unit 14 generates a predicted residual image by subtracting the predicted image corresponding to the image of the processing target block in pixel units from the image (input image) of the processing target block (step S5). This predicted residual image is output to the conversion unit 150. Furthermore, the conversion unit 15 converts the predicted residual image generated by the subtraction unit 14 using a specific conversion rule, and calculates a conversion coefficient obtained by the conversion. (Conversion Information) (step S6). This conversion coefficient is output to the encoding section 16 and the inverse conversion section 17. ☆ Next, the encoding unit 16 encodes the conversion coefficient calculated by the conversion unit 15 based on the specific rule (step S7). In addition, the encoding unit 16 encodes the coding mode information determined in step S2 on the basis of a specific rule, and, based on the specific rule, generates a pattern based on the prediction information selected in the average information amount encoding step S4. The average amount of information encodes these encoded data, produces compressed data (bit stream), and outputs -29- 200529673 (26) to an external image decoding device. · Secondly, the “inverse conversion unit 17” adopts the inverse conversion rule corresponding to the specific conversion rule adopted by the conversion unit 15 and the conversion coefficient calculated by the inverse conversion conversion unit 15 to generate a recovery prediction residual image (step S8). This reply prediction residual image is output to the adding unit 18. Furthermore, the adding unit 18 adds the predicted picture generated by the predicted picture generating unit 13 and the reply predicted residual picture generated by the inverse conversion unit 17 to generate a playback picture (step S9). The playback image is stored in the memory through the storage unit 19 and stored (step S 1 0). Next, it is judged whether or not the processing is completed for all the blocks (step S1 1). When all the blocks are completed (steps, steps S 1 1; YES), the image encoding processing is ended. On the other hand, if all the blocks have not been completed (step S11; NO), the process proceeds to step S2. Next, an image encoding program related to the present invention and a computer-readable recording medium (hereinafter referred to as a recording medium) recording the image encoding program will be described. The recording medium mentioned here refers to the reading device provided for the hardware resources of the computer. In response to the description content of the program, the state of changes in energy such as magnetism, light, electricity, etc. can be caused by the corresponding signal form. The description of the program is transmitted to the reading device. Relevant recording media include, for example, magnetic disks, optical discs, CD-ROMs, and megabytes built into the computer. 0 'FIG. 8 is a diagram showing the structure of a recording medium related to the first embodiment. As shown in FIG. 8, the recording medium 100 has a program area 101 of a recording program. In this program area 1 0 1 records an image coding program 10 2. -30- 200529673 (27) Fig. 14 is a diagram showing a hardware structure of a computer for executing a program recorded on a recording medium, and Fig. 15 is a perspective view of a computer for executing a program recorded on a recording medium. As shown in FIG. 15, the computer 1 10 includes a reading device 1 1 2 such as a floppy disk drive device, a CD-ROM drive device, and a dvd drive device, and a working memory (ram) 1 in which an operating system usually exists. 1 4 and memory 1 1 for programs stored in the recording medium 1 100, and display devices 118 such as displays, mouse 120 and keyboard 122 for input devices, and communication devices for receiving and transmitting data, etc. 124, and CPU 126 which controls the execution program. If the recording medium 100 is inserted into the reading device 112, the computer 1 10 will be able to access the image encoding program 102 stored in the recording medium 100 from the reading device 112, and the portrait encoding program 102 may be used. It is assumed that the image encoding device 1 operates. As shown in Fig. 15 ', the image encoding program 102 can also be provided as a computer data signal 130 which is superimposed on a carrier wave and is provided via a network. At this time, the computer 110 can store the image encoding program 102 received by the communication device 124 into the memory 116, and execute the image encoding program 102. The image encoding program 102 includes an image segmentation module 102a, an encoding mode determination module 102b, a predictive image generation module 102c, a subtraction module 102d, a conversion module 102e, an encoding module 102f, and an inverse conversion module. The group i02g, the addition module 102h, and the memory module 102i are configured. Here, the image segmentation module 102a, the encoding mode determination module 102b, the predictive image generation module 102c, the subtraction module 10, the conversion module 102e, the encoding module 102f, and the inverse conversion module are used. Group 102g, addition module 102h and memory module-31-200529673 (28) Function implemented by each operation of '102i' and the image division unit 1 of the image encoding device 1 described above 1, encoding mode determination unit 1 2, The predicted image generation unit 1 3, the subtraction unit 14, the conversion unit 15, the encoding unit 16, the inverse conversion unit 17, the addition unit 18, and the memory unit 19 each have the same function. If the image encoding device of the first implementation type is used, "for the field where the encoding mode is the complement coding mode, it is not necessary to include the data generated based on the prediction assistance information in the bit stream, so it has a high encoding efficiency. Yuan flow. Also, the predicted image generation unit 13 can generate a prediction by using a playback image of a non-connected block that is not connected to the processing target block when the coding mode of the connected block connected to the processing target block is the complement coding mode. portrait. As a result, because not only the playback portrait of the connected block connected to the processing target block, but also the playback portrait of the non-connected block that is not connected to the processing target block, the reference range when generating the predicted portrait is expanded. , And can reduce the redundant length in space, and at the same time can improve the efficiency of quotient coding. In addition, by more effectively restricting the picture of the complement coding mode, the reference range when generating the predicted picture will be more effectively expanded, and the redundant length in space will be reduced. < [Second Embodiment Mode] Next, a second embodiment mode of the present invention will be described. This portrait decoding device receives compressed data (including encoded data>), that is, a bit stream, output from the portrait encoding device of the first embodiment, and decodes the received bit stream to generate playback portrait data. Figure 9 shows Fig. 32 shows the structure of an image decoding device related to the second embodiment. (29) (29) 200529673. The image decoding device 2 shown in Fig. 9 can be physically equipped with, for example, a CPU (Central Processing Unit) and memory. Memory devices such as personal computers, display devices such as monitors, and computers for communication devices. The image decoding device 2 can also be a mobile communication terminal such as a mobile phone or a DVD device. That is, the image decoding device 2 can be widely applied to devices capable of information processing. The image decoding device 2 shown in FIG. 9 includes a decoding unit (decoding means) 2 1. An encoding mode determination unit 2 2. A predicted image generation unit (the first image prediction means) 23. An inverse conversion unit (inverse conversion means) 24. Adding section (playback image generation means) 25, memory section (memory means) 26, switch 27, and supplementary image generation section (second image prediction means) 28. Next, for Each of the structural elements shown in 9 will be described. The decoding unit 21 receives the input image information (compressed data) of the input image of the decoding processing object divided into blocks of a specific size. The decoding unit 21 averages the amount of information based on a specific rule. Decode the received input image information. Decode the conversion coefficient, encoding mode information, and prediction mode information by decoding the average information amount. These conversion coefficients, encoding mode information, and prediction mode information are different from the description about the image encoding device 1 The conversion coefficients, encoding mode information, and prediction mode information of the quantitative image data are the same, so description is omitted. The encoding judging unit 22 judges the encoding mode from the input encoding mode information according to a specific scanning order (such as a light spot scanning order). It is either a predictive coding processing mode or a complement coding mode. The processing performed by the coding judgment unit 22 here is different depending on whether the scan in the frame is the first round or the second round. Specifically, when the scan in the frame is the first round, it is judged that the encoding mode is predictive encoding processing-33- 200529673 (30) mode In the case of the formula, the predicted image decoding process including the first image prediction process described above is performed on the processing target block; when the encoding mode is determined to be a complement coding mode, the next block is read. On the other hand, the scan in the frame For the second round, determine whether the encoding mode is the complement coding mode, and perform the complement image decoding processing including the above-mentioned image complement processing on the processing target block; if the encoding mode is the predictive encoding processing mode, read Take the next block. That is, in the first round, only the prediction target decoding process including the predictive image decoding process is performed on the processing target block of the predictive encoding processing mode, and in the first round only the subtraction encoding processing mode is processed. The block performs the supplementary image decoding process including the supplementary image process. The prediction image generation unit 23 completely decodes the image of each block in accordance with the prediction mode limited by the decoded prediction mode information, and uses a part of the playback image recorded by the accounting unit 26 billion to generate a processing object to be decoded. Predicted portrait of a block. The image prediction process for generating this predicted image is the same as the first image prediction process described above (refer to FIG. 2 and FIG. 3), and therefore description is omitted. The inverse conversion section 24 inverse-transforms the decoded conversion coefficients using an inverse conversion rule corresponding to the specific conversion rules adopted by the above-mentioned conversion section 15 and generates a residual prediction residual image obtained by the inverse conversion. The adding unit 25 adds the predicted image and the response corresponding to the predicted image to predict the residual image to generate a broadcast image. In addition, when the addition unit 25 sets a specific range of the picture element 値 of the image, the picture element 値 may be limited to a certain range and subjected to clipping. The memory 26 stores the playback image generated by the addition unit 25 in the memory shown in the figure below -34- 200529673 (31). The changeover switch 27 switches the transfer position of the playback image stored in the storage unit 26 in accordance with the encoding mode of the processing target block. That is, the switching switch 2 7 is switched when the encoding mode is the predictive encoding processing mode, so that the playback image stored in the memory section 26 can be transmitted to the predictive image generating section 23. On the other hand, when the encoding mode is the subsequence coding processing mode, the switch is switched so that the playback image stored in the storage unit 26 can be transferred to the subscribing image generation unit 28. . The supplementary image generation unit 28 generates a supplementary image by playing the decoded image in the vicinity of the processing target block. Here, the image subtraction processing when generating a sub-image is the same as the above-described image sub-processing (refer to FIG. 6 and the second image prediction processing), so description thereof is omitted. Next, operations of the image decoding device 2 and image decoding processing will be described with reference to FIGS. 10 to 12. First, the operation of the image decoding process will be described with reference to FIG. 10. As shown in FIG. 10, in the image decoding process, first, the decoding unit 21 receives the input image sales information of the 1 frame portion from the image encoding device 1 of the first embodiment type to decode based on the average information amount of a specific rule To generate conversion coefficients, encoding mode information, and prediction mode information (step S20). Next, perform the prediction picture decoding process performed in the frame scan in the first round (step S30), and then perform the second round The supplementary image decoding process (step S40) performed by the cat in the frame. Hereinafter, each operation of the predictive image decoding process (step S30) and the supplementary image decoding process (step S40) will be described in detail in each process. 35- 200529673 (32) First, referring to Fig. 11, the explanation of the predicted portrait will be explained. First, the encoding judging unit 22 judges whether the input encoding code mode is the prediction encoding processing mode according to, for example, the scanning order of light spots. (If the step is NO (step S3 1; NO), the process is shifted. On the other hand, in the judgment of step S31, the case of the encoding processing mode is judged. (Step 31; YES)) The unit 23 follows the above. In the prediction mode defined by the information in step S20 (refer to FIG. 10), the pictures of each block are stored in a part of the play image of the play image in the memory section 26 (step S32). Furthermore, the production method is based on the above-mentioned 1 Image prediction processing. Also, in the addition section 25. Second, the inverse conversion section 24 uses the inverse conversion rule corresponding to the specific conversion rule adopted above with respect to the conversion coefficient decoded in step 10) above, and applies the inverse conversion rule. The converted predicted residual image obtained by the conversion This residual predicted residual image is output to the adding section 25. Next, the adding section 25 is the addition of the predicted image production and the predicted image, and the reverse conversion of the reverse conversion 24 to generate a playback image. The image is memorized and stored in memory (step S35). Next, it is judged whether the processing ends the detailed scanning sequence of the code processing for all the blocks (the information is limited to the compilation S 3 1). This judgment is set in the following steps S 3 6 The encoding mode is prediction. The prediction image is completely decoded in the prediction mode image generated by decoding. Output test illustration: step S20 (refer to FIG. 1 conversion unit 5 inverse transform, and generates (step S 3 3) 23 generates three portraits of the predicted residual. The unit 26 is stored in the record (step S36), -36- 200529673 (33) When all the blocks are completed (step S36; YES), the prediction image decoding process is ended. On the other hand, if all the blocks are not completed (step S36; NO), the process shifts to the above step S31. The detailed operation of the sub-picture decoding process (step S40 in FIG. 10) will be described with reference to FIG. First, the encoding judging section 22 determines whether the encoding mode defined by the inputted encoding information is a complement coding mode according to a specific scanning order (for example, a light spot scanning order) (step S4 1). When this determination is NO (step S41; NO), the processing moves to step S44 described later. On the other hand, in the judgment of step S41, it is judged that the encoding mode is the complement coding mode (step 41; YES). The complement image generating unit 28, for each of the processing target pixels included in the processing target block, starts from Playback images located around each processing target pixel, obtain the playback pixel 値 with the greatest correlation, calculate 塡 supplemental pixel 値, thereby generating a supplementary portrait corresponding to the processing target block (step S42). In addition, the method for generating this supplementary portrait is based on the aforementioned portrait supplementation processing (second portrait prediction processing). Next, the supplementary image generated by the supplementary image generation unit 28 is used as a playback image, and is stored in the memory by the storage unit 26 and stored (step 543). Next, it is determined whether the processing is completed for all the blocks (step S44). ), When all the blocks are finished (step S44; YES), the sub-picture decoding process is ended. On the other hand, if all the blocks are not completed (step 544; NO), the process proceeds to step S41 described above. Fig. 13 is a diagram showing the structure of a recording medium related to the second embodiment type -37- 200529673 (34). As shown in FIG. 13, the recording medium 100 includes a program field 201 for recording a program. In this program field 201, the image decoding program 202 is recorded. ° If the recording medium 100 is inserted into the reading device 112, the computer 110 (refer to FIGS. 14 and 15) can access the image decoding program 202 stored in the recording medium 100 from the reading device 112 The image decoding program 202 can be set to operate the image decoding device 2. As shown in FIG. 15, the image decoding program 202 can also be provided as a computer data signal 130 which is superimposed on a carrier wave and is provided via a network. At this time, the computer 110 may store the image decoding program 202 received by the communication device 124 into the memory 116 and execute the image decoding program 202. The image decoding program 202 includes a decoding module 202a, an encoding mode determination module 202b, a predictive image generation module 202c, an inverse conversion module 202d, an addition module 202e, a memory module 202f, a switch module 202b, and 塡Compensation image generation module 202h. Here, the decoding module 202a, the encoding mode determination module 202b, the predictive image generation module 202c, the inverse conversion module 202d, the addition module 202e, the memory module 202f, the switch module 202g, and 塡The functions implemented by the respective operations of the supplementary image generation module 202h are the same as those of the decoding unit 2 1, the decoding mode judgment unit 22, the predicted image generation unit 23, the inverse conversion unit 24, the addition unit 2 5, and the memory unit 1 of the aforementioned image decoding device 2. 6. Switch 2 7. Each of the supplementary image generating units 28 has the same function. If the image decoding device 2 of the first embodiment is used, it can effectively decode the image encoded by the image encoding device 1. That is, the prediction image generation unit 23 can use a non-connected block not connected to the processing target block when the coding mode of the connected block connected to the processing target block is 塡 -38- 200529673 (35). Play the portrait to generate a predicted portrait. As a result, because not only the playback portrait of the connected block connected to the processing target block, but also the playback portrait of the non-connected block that is not connected to the processing target block, the reference range when generating the predicted portrait is expanded. And reduce the length of space. In addition, when it is determined that the encoding mode is the subsequence coding mode, a fully decoded playback image may be used to generate a supplemental portrait that complements the playback image. Therefore, even if the image in the supplementary coding mode is not transmitted during encoding, it is possible to generate a playback image. In addition, in each of the above-mentioned embodiments, although the in-frame coding process has been described as the image prediction of the predicted image generating units 1 3 and 23, it is not limited to this. For example, motion compensation prediction may be included in Portrait prediction. In this case, it is also possible to encode the motion vector information with an average amount of information. In addition, as this motion compensation prediction, the same method as the motion compensation prediction of MPEG-4 or H.264 which has been used conventionally can be adopted. It is also possible to apply various motion compensation prediction methods. Also, in each of the above embodiments, H. is used.  The prediction of the spatial domain used in 264 is used as the in-frame prediction, but the in-frame prediction is not limited to this. For example, the prediction in the frequency domain by MPEG-4 can also be applied. At this time, the above or left connected block encoding or decoded case uses all or part of the connected block conversion coefficients as the encoding. Or predict the conversion coefficient of the target block. If the upper and left connected blocks are both coded or decoded, it shows which of the blocks connected by the upper and left -39- 200529673 (36) edges. Bit data that is encoded as prediction mode information. In addition, in the case where any one of the blocks connected above or to the left is coded or decoded, all or a part of the orthogonal conversion coefficients of the coded or decoded block are used as the prediction frame. On the other hand, in the case where the encoding or decoding of any of the blocks connected above or to the left is not completed, the orthogonal conversion coefficients of all the encoded or decoded blocks located at the distance closest to or above the block are used. Or part of it, as a prediction. However, if the coded or decoded block located closest to the upper and left distance is located at the same distance as the upper and left, the upper and left block shows the conversion coefficient of which block is used. Or a part of it is used as 1-bit data of the prediction frame, and it is encoded as prediction mode information. In addition, if all the blocks above and to the left of the block to be coded or decoded are coded or decoded incompletely, only the DC component is set to 値 (e.g., 1 28) for prediction. In each of the above embodiments, the prediction frame can be set to 0 by in-frame prediction, and the operation can be performed without performing in-frame prediction. In this case, since the redundant length of the space is reduced by introducing the portrait complement mode, the encoding efficiency can also be improved. In each of the above-mentioned embodiments, although the portrait playback process uses a playback image in the same space as the processing target frame to generate a predicted image, it can also be used in a frame different from the processing target frame. The playback image of the frame that generates the playback image at the end is used for the image complementing process. Also, in the above first implementation type, • 40-200529673 (37) produced by the image correction process is still used as the playback image, but in terms of encoding, the predicted image and the predicted residual image of the input image are also The above conversion processing and average message encoding can be performed and included in the bit stream. In this case, in the decoding aspect, the decoded and inversely processed bit stream data is used to generate a restored predicted residual image, and the restored predicted residual image is added by adding the predicted residual image and the predicted image generated by the image complementation process to generate Play portrait. [Third Embodiment] Hereinafter, a third embodiment of the present invention will be described. Figure 16 • Each figure shows the structure of an animation encoding device related to the third embodiment of the present invention. The animation encoding device 3 10 shown in FIG. 16 can be physically configured as, for example, a computer having a CPU (Central Processing Unit), a memory device such as a memory, a display device such as a display, and a communication device. The moving picture encoding device 310 may be a mobile communication terminal such as a mobile phone or a DVD device. That is, the image encoding device 310 can be widely applied to devices capable of information processing. The moving picture encoding device 310 is functionally provided with a domain division unit 312, an encoding mode determination unit (determination means) 3 1 4, an image prediction unit 3 1 6, a subtraction unit 318, a conversion unit (conversion means) 320, and an encoding unit (encoding means) 322. 1. Inverse conversion unit (inverse conversion means) 3 24. Addition unit (playback image generation means) 3 2 6 and image memory unit (memory means) 3 2 8. The field segmentation section 312 sequentially sets each frame of the input image (animation) as a frame to be encoded, and divides the frame to be encoded into a plurality of partial regions (hereinafter referred to as microblocks) of a specific size. The specific size may be, for example, a size of 16 × 16 pixels, but is not limited thereto, and may be other sizes. -41-(38) (38) 200529673 The domain segmentation section 312 generates the microblock position information for limiting the microblock. For example, the location information of the microblocks can be an integer number marked from large to small according to the scanning order of the light spots in each microblock in the frame. In addition, the position information of the microblocks can use the pixel coordinates located at the front of the microblocks according to the scanning order of the light spots. The encoding mode determination unit 314 obtains the encoding mode of each microblock, and outputs the encoding mode information that restricts the encoding mode to the portrait prediction unit 316 and the encoding unit 322. In this implementation form, the coding mode information is used to limit the information that should be used to generate the predicted image of the microblock by the first image prediction process or the second image prediction process. In the following, the first image prediction process is set to motion compensation prediction, and this embodiment will be described. In the description below, the encoding mode when the micro-block is processed by the first image prediction process is set to the P mode, and the encoding mode when the micro-block is processed by the second image prediction process is set to the C mode. The first image prediction process may be, for example, a known Internet frame prediction process. Details of the first image prediction process and the second image prediction process will be described later. The image predicting unit 3 1 6 generates a predicted image of a microblock to be processed based on the encoding mode information output by the encoding mode determining unit 3 1 4 using either the first image predicting process or the second image predicting process. The image prediction unit 3 1 6 processes the microblocks in the first image prediction process, and outputs the motion vector of the prediction assistance information to the encoding unit 322. Details of the portrait prediction section 316 will be described later.丨 The subtraction unit 3 1 8 performs the prediction image of the processing target microblock generated by the image prediction unit 3 1 6 and the processing pair output from the domain segmentation unit 3 1 2 -42- 200529673 (39) Image microblock The difference between the input portraits is calculated to produce a predicted residual portrait. The conversion unit 320 converts the predicted residual image and generates conversion information. For this conversion, conversion processing and dequantization processing such as DCT (Discrete Cosine Transform) can be used. In this case, the conversion information includes a set of quantization coefficients. The DCT can be a 2-dimensional element DCT with 4 rows and 4 columns, or a 2-dimensional element DCT with 8 rows and 8 歹 ij. In addition, this conversion can use arbitrary conversion processing such as 4 rows and 4 columns of integer straight-line conversion and quantization, MP method (Matiching Pursuit), or vector quantization and wavelet transform (Quantum) used in H · 264.化 处理。 Processing. The encoding unit 3 22 encodes the microblock position information from the domain segmentation unit 312, the encoding mode information from the encoding mode determination unit 314, and the motion vector from the image prediction unit 3 1 6 and the conversion from the average information amount. The encoding processing such as conversion information of the unit 320 is encoded, and the encoding result is included in the bit stream and output. The anti-conversion section 324 generates a re-predicted residual image by re-converting the conversion information from the transform section 320 to return the predicted residual image. In the case where the conversion information includes the quantization coefficient set, the inverse conversion unit 324 performs an inverse quantization process and an inverse conversion process on the quantized coefficients, and generates a return prediction residual image. The adding section 3 2 6 generates a playback image by adding the predicted predicted residual image from the inverse conversion section 3 2 4 and the predicted image from the image prediction section 3 1 6 to the playback image frame, which is then stored in the playback frame. Portrait memory 328. In addition, in the case where the range of the picture element 画像 of the image data is set, the addition unit 326 may also capture the picture element 结果 of the addition result in order to condense within the range. -43- 200529673 (40) Recording 100 million copies of 3, 2 8 memories, complete the micro-blocks of the encoding target frame, and play the playback frame and reference frame of the portrait. This reference frame is a playback frame that is different from the encoding object. In this embodiment, it is a playback frame before a frame. The encoding mode determination unit 3 1 4 will be described below. Fig. 17 is a diagram for explaining the processing of the encoding mode determining section. Regarding Fig. 17, the area of the quadrangle is represented as a block, and the solid line and dotted line marked with arrows indicate the scanning order. As shown in (a) of this figure, the encoding mode determination unit 314 firstly generates a first image prediction process (motion compensation prediction process) in a specific scanning order (for example, a light spot scanning order) by encoding the target frame The predicted pictures of all the microblocks (that is, the predicted pictures are generated in the P mode), and the playback picture frame formed by the playback pictures of all the microblocks of the coding target frame are generated. The encoding mode decision unit 314 is as shown in this figure (b). As shown, in the scanning order opposite to the above, by generating a pre-drawn portrait of all the microblocks of the coding mode object frame of the second picture prediction process, a playback portrait of all the microblocks of the coding object frame is generated. 1 Encoding mode determination unit 3 1 4 When the second image prediction process is used to generate the predicted image, the micro-blocks of the lithium object are 'adopted to play the image in the micro-block located in a specific scanning order' The scanning order is located in the rear microblock, the playback image of the microblock with the encoding mode 1>, and the reference frame recorded in the image memory 328. For example, as shown in (c) of this figure, if the micro-block marked "A" is set as the micro-region of the processing object -44- (41) (41) 200529673, the micro-block of the playback frame will be used. , Compared to the playback portrait of the microblock in front of block A, compared to the playback portrait of the microblock behind A and the encoding mode is P mode (microblock labeled "P" in the figure), and The reference frame stored in the image storage section 3 28 performs the second image prediction process. The details related to the second image prediction process will be described later. Each time the encoding mode determination unit 3 1 4 generates a playback image after the second image prediction process, it calculates the costs associated with the playback image generated by the second image prediction process and the playback image generated by the second image prediction process. Decide the encoding mode. This cost is, for example, the number of bits R required for encoding the microblock of the processing target, the square D of the error between each pixel of the playback image of the microblock and each image of the input image of the microblock, and a preset value. When the coefficient is set to λ, D + λ and R are calculated. The encoding mode determination unit 314 selects the P mode or the C mode, and the mode with the lower cost is set to the encoding mode. In addition, the cost can be calculated from various functions if it can show coding efficiency or / and portraits.  The image prediction unit 316 will be described in detail below. Fig. 18 is a block diagram showing the structure of the image predicting section 3 1 6. The image prediction unit 3 1 6 includes a mode switching unit 3 3 0, a first image prediction unit (first image prediction means) 3 32, and a second image prediction unit (second image prediction means) 3 3 4. The mode switching unit 330 receives the encoding mode information from the encoding mode determining unit 314, and activates the first image prediction unit 3 3 2 or the second image prediction unit 3 34 according to the encoding mode defined by the encoding mode information. 34 The mode switching unit 3 3 0, when the encoding mode is P mode, the first day image prediction unit -45- 200529673 (42) 3 3 2 is activated, and when the encoding mode is C mode, the second image prediction unit 3 3 4 is activated. In the animation encoding device 310, the 'mode switching unit 330' first generates the predicted portraits of all the micro-blocks in the encoding target frame by using the first image prediction unit 3 3 2 and the encoding mode is the P-mode. Thereafter, the mode switching unit 3 3 0 generates predicted pictures of all the micro-blocks in the coding object frame with the second picture prediction unit 3 3 4 in the coding mode of the C-mode. The first image prediction unit 332, 'as described above, generates a playback image by the first image prediction process, that is, the motion compensation prediction process. The so-called motion compensation prediction processing is to match the input image of the target microblock with the block matching in any area of the reference frame, and use the part of the image with the highest correlation in the reference frame area as the playback image. Processing of moving vectors to the field. This relationship uses, for example, the square D of the difference between the input image of the microblock processed by the object and the playback image of the matched object's field, the data length R required for encoding the motion vector from the processed microblock to the field, and The predetermined coefficient λ is calculated from a reference function of D + λ R. In addition, the motion compensation prediction process can also be performed by using an image with upsampling reference frames of 2 or 4 times. In addition, the motion compensation prediction process may be performed on a microblock divided into smaller blocks. At this time, a block partitioning pattern indicating the movement vector of each small block and the type of microblock partitioning will be output. The second image prediction unit 3 3 4 uses the second image prediction processing to generate a predicted image of a microblock whose encoding mode is C mode. Fig. 19 is a diagram for explaining the second image prediction processing -46- 200529673 (43). In the second image prediction process, 'the microblock Y where no image is played is selected. In this micro-block Y, an unprocessed pixel ′ of a picture element 预测 (prediction signal) to which an image is played is selected, and a template T is set in which this picture element is included. In this implementation form, although the template T mainly composed of the above-mentioned unprocessed pixels is set, if the processed and unprocessed pixels are included, the template T can be of any size and shape. For the microblock γ, a reference area R of an arbitrary size is set relative to the playback frame F1. Also, for the reference stored in the portrait memory 328. In frame F2, the reference area R is also set. Reference collar for frame F2 •. The field R may also be set at a position corresponding to the reference frame F2 set at the position of the reference field R of the playback frame F1. Alternatively, the reference area R of the reference frame F2 may also be set at a position deviated from the motion vector generated by median in the motion vector corresponding to the microblock around the microblock Y. In the second image prediction process, the candidate field S of the generated playback image located at an arbitrary position in the reference field R is calculated to correlate with the template T, and the candidate field with the highest correlation selects the copy reference field Sc. This correlation is for example modulo. The corresponding positions of the board T and the candidate field S generate pixels for playing the picture pixels ，, which are set as effective pixels. The square of the difference 値 between the template T and the pixels 有效 of the effective pixels in the candidate field S, that is, The matching error M is calculated from 値 divided by the effective pixel number U. In addition, if the correlation can present the pixel similarity of the template T and the candidate field S, it can also be calculated by various calculation methods. In the second image prediction process, on the unprocessed pixels in the template T that have not been played with the image, the corresponding reference field Sc is copied-47- 200529673 (44) The pixels of the pixel are used as塡 make up pixels 値. In FIG. 19, only the part where the playback image of the template T has been generated is marked with a diagonal line. In Fig. 19, "the area in which the playback image is not generated in the template D (the lower part of the template D)" is presented and copied to the lower reference pixel SC. In the second image prediction process, the relevant microblocks are repeatedly selected up to the unprocessed pixels of the picture element to which the picture is played, and the picture is repeatedly selected. The operation of the moving picture encoding device 3 10 will be described below. At the same time, an animation coding method related to the implementation form of the present invention will be described. FIG. 20 is a flowchart showing a method of animation coding related to the third embodiment. As shown in FIG. 20, in the encoding process of the animation encoding device 310, the encoding target frame is divided into a plurality of microblocks by the domain segmentation unit 312 (step S301). The domain segmentation unit 312 generates the above-mentioned microblock position information. Next, an encoding mode determination process for determining the encoding mode of each microblock by the encoding mode determination unit 314 is performed (step S302). Fig. 21 is a flowchart showing a coding mode decision process of the animation coding method according to the third embodiment. As shown in FIG. 21, in the encoding mode determination process, first, according to the scanning order (light spot scanning order), the first image prediction process (motion compensation prediction process) is used to generate the playback images of all microblocks. Play frame (step S302-1). Then, follow the scan order. The last microblock is the processing target microblock (step S3 02-2), and the second image prediction process described above is executed. In the second image prediction process. As shown above, the play block generated by step S3 02-1 is used to scan the microblock of the processing object.  -48- 200529673 (45) The playback image of the micro area & the scan order is located in the front, and the micro block of the processing object is located in the micro block in the scan order. The encoding mode is P_micro, block The playback image and the reference frame to generate the playback image. Second, the cost of the playback image of the processing target microblock generated in step S3 02-3 and the processing of the playback image of the processing target microblock generated in step S3 02-1 are based on the cost function described above. Calculation (step S3 02-4). Next, in the first portrait prediction process, that is, the P mode, or the second portrait prediction process, that is, the C mode, a lower cost mode is selected as the encoding mode (step S302-5). % Furthermore, it is tested whether the processing of all microblocks is completed (step S302-6). When all microblock processing ends (Yes), the encoding mode is ended and the processing is determined. On the other hand, in the case that all microblock processing is not completed (No), in accordance with the reverse scanning order, a microblock with an undecided encoding mode is selected (step S302-7), and steps S302-3 to S302 are repeated. -6 processing. Returning to Fig. 20, in the animation encoding device 310, encoding mode information is encoded by the encoding unit 322 (step S303). Then, the mode switching unit 330 selects a microblock as a processing object according to a specific scanning order (for example, a light spot scanning order), and tests whether the coding mode of the processing microblock is P mode (first image (Prediction processing) (step S304). When the encoding mode is not P mode (No), the manufacturing process moves to step S312. On the other hand, when the encoding mode is P mode (Yes), the mode switching unit 3 3 0 activates the first image prediction unit 3 3 2 and generates a predicted image of the microblock to be processed by the first image prediction process (step S3 05 -49- 200529673 (46)) ° Furthermore, the subtraction unit 318 generates a predicted residual image formed by the difference between the input image and the predicted image of the processing target microblock (step S3 06). Next, the predicted residual image is converted into conversion information by the conversion unit 320 (step S307). Then, the motion vector and the conversion information are encoded by the encoding unit 322 with the average information amount (step S 3 0 8). Next, inverse conversion is performed on the conversion information by the inverse conversion section 324, and a return prediction residual image is generated to return the prediction residual image (step S3 09). Furthermore, the addition residual unit 326 adds the predicted residual image and the predicted image from the image prediction unit 3 16 to generate a playback image of the processing target microblock (step S3 1 0). The playback picture is grouped into a playback frame and stored in the picture memory 3 2 8 (step S 3 1 1). Furthermore, it is tested whether the processing of all the microblocks is completed (step S312). In the case where the processing of all the microblocks is not completed (No), the unprocessed microblock is selected (step S302-7), and the processing from step S304 to step S312 is repeatedly performed. On the other hand, if all microblock processing is completed (Yes), the process moves to step S313. In step S3 13, the mode switching unit 3 3 0 selects a microblock as a processing object according to a specific scanning order (for example, a light spot scanning order), and tests whether the coding mode of the processing microblock is C mode (second image prediction processing). When the coding mode is not the C mode (No), the process moves to step S312. On the other hand, when the encoding mode is C mode (Yes), the mode switching unit 3 3 0 activates the second image prediction unit 3 34 and processes the second image prediction -50- 200529673 (47) to generate a micro target. The predicted portrait of the block (step S3 1 4). Fig. 22 is a flowchart showing a second picture prediction process of the animation coding method according to the third embodiment. As shown in Fig. 22, in the second image prediction process, a block in which a playback image is not generated is selected. In this embodiment, the microblocks whose coding mode is C mode are set as the microblocks to be processed, and are selected according to a specific scanning order (step S3 14-1). Next, as described above, the reference area for the microblock to be processed is set (step S314-2). Next, select a pixel in which no playback image is generated in the microblock (step S314-3), and set a template that includes this pixel in a part of it (step S3 14-4). As shown above, the correlation between this template and the candidate fields of the reference field is calculated (step S3 14-5), and the candidate field with the highest correlation is selected as the copy reference field (step S314-6). Secondly, copy the unprocessed pixels in the template to which the picture element 値 of the image is not played, and copy the picture element 値 of the corresponding pixel in the reference area (step S3 14-7). Then, in the micro block of the test processing object, Are there any unprocessed pixels that have not been played back? (Step S314-8). If there is an unprocessed pixel (Yes), select the unprocessed pixel (step S314-9), and repeat steps S314-4 to S314-8. On the other hand, if there is no unprocessed pixel in the processing target microblock, the second image prediction process is terminated (step S314). Returning to FIG. 20, in the animation encoding device 310, the subtraction unit 318 generates a difference between the input image of the processing target microblock and the second image prediction section-51-200529673 (48). The predicted residual image formed (step S315). ~ Next, the predicted residual image is converted into conversion information by the conversion unit 320 (step S3 16). Then, the encoding unit 322 averages the amount of information to encode the conversion information (step S3 17). The inverse conversion section 324 performs inverse conversion on the conversion information, and generates a return prediction residual image that returns the predicted residual image (step S3 1 8). In addition, by adding an addition section 3 2 6, the restored prediction residual image and the predicted image from the image prediction section 3 1 6 are added to generate a playback image of the processing target microblock (step S3 1 9 >. Playback The portraits are grouped into the play frame and stored in the portrait memory 3 2 8 (step S 3 2 0). Furthermore, it is tested whether the processing of all micro-blocks is completed (step S 3 1 2). The processing of all micro-blocks is not completed. (No), select ^ unprocessed microblocks (step S3 02-7), and repeat the processing from step S3 04 to step S312. On the other hand, all microblock processing ends (Yes) The encoding process will be ended. In the following, the animation encoding program that sets the computer as an animation encoding device 3 10 will be described. Fig. 23 shows the structure of the animation encoding program related to the third embodiment and the recording media. The animation coding program 340 shown in FIG. 23 is stored in the recording medium 100 for use. The recording medium 100 is, for example, a recording medium such as a magnetic disk, CD-ROM, DVD, or ROM, or a semiconductor memory. Media 100 Into the reading device 112, the computer 110 (refer to FIG. 14 and FIG. 15) will be able to access the animation encoding program 340 stored in the recording medium-52-200529673 (49) 100 from the reading device 112, and may encode by the animation The program 340 operates as the animation encoding device 310. As shown in FIG. 15, the animation encoding program 3 40 can also be provided as a computer data signal 130 that is superimposed on a carrier wave, and provided through the network. At this time, the computer 1 10 can The animation encoding program 3 40 received by the communication device 124 is stored in the memory 116, and the animation encoding program 340 is executed. As shown in FIG. 23, the animation encoding program 3 4 〇 is provided with a main module 341 and a field with overall processing. Segmentation module 342, encoding mode determination module 344, portrait prediction module 346, subtraction module 348, conversion module 3 50, encoding module 3 5 2, inverse conversion module 3 54, addition module 3 56 and Image memory module 3 5 8. The image prediction module 346 includes a mode switching module 3 60, a first image prediction module 3 62, and a second image prediction module 364. A domain segmentation module 342 and a coding mode determination module 344, portrait prediction module 346, subtraction module 348, turn Module 3 50, encoding module 352, reverse conversion module 3 5 4, addition module 3 5 6, image memory module 3 5 8, mode switching module 360, first image prediction module 362 and second image The prediction modules 3 64 each execute the functions of a computer, and the above-mentioned domain division section 3 1 2, encoding mode determination section 3 1 4, image prediction section 3 1 6, subtraction section 3 1 8, conversion section 320, encoding section 322, The inverse conversion unit 324, the addition unit 326, and the portrait memory unit 328, the mode switching unit 330, the first image prediction unit 332, and the second image prediction unit 334 are the same. In the following, the functions and effects of the animation encoding device 3 10 will be described. If the image encoding device 3 10 of the third embodiment is used, the encoding mode is C-mode micro-blocks, that is, predictions generated by the second image prediction process-53- 200529673 (50) micro-blocks of images, Since it is not necessary to include the data generated based on the prediction assistance information in the bit stream, a bit stream with high coding efficiency is generated. In the moving picture encoding device 310, the second picture prediction unit 3 34 generates a predicted picture by using the reference picture frame and the playback picture frame of the encoding target zombie frame. This play frame includes a play image completely generated through the first picture prediction process and the second picture prediction process. That is, the second image prediction unit 334 generates a predicted image by using a reference frame formed by a playback image of a frame different from the encoding target frame and a playback frame formed by a playback item of the encoding target frame. . As a result, coded data can be produced which can reduce the verbosity of the time direction and the space direction.丨 In the second image prediction process, a play frame including a playback image completely generated through the first image prediction process and the second image prediction process is used. Therefore, it is also possible to use a playback image located behind due to the scanning order for prediction, so that the redundant length in the spatial direction can be effectively reduced. [Fourth embodiment] The following is a description of an animation decoding device according to a fourth embodiment of the present invention. Fig. 24 is a diagram showing the structure of an animation decoding device according to a fourth embodiment of the present invention. The animation decoding device 3 70 shown in FIG. 24 can be physically a computer equipped with a memory device such as a CPU (Central Processing Unit), a memory, a display device such as a display, and a communication device. The animation decoding device 3 70 may be a mobile communication terminal such as a mobile phone or a DVD device. That is, the animation decoding device 3 70 can be widely used as an information processing device-54- 200529673 (51) As shown in FIG. 24, the animation decoding device 3 70 has a solution means) 372 and an inverse conversion unit (inverse conversion means) 3 74 , 3 76, the addition department (playing portrait generation means) 3 78 and painting memory means) 3 80. -The decoding unit 3 72 receives the input bit stream and decodes the bit stream 372 which receives the bit stream generated by the animation encoding device 310 to decode the bit stream to generate microblock location information, motion vectors, and conversion. Information. The inverse conversion unit 3 74 receives the conversion information from the decoding unit 3 72 and performs inverse conversion on the micro-block conversion information of the object, thereby generating a micro-block response prediction residual image. The inverse conversion unit 374 outputs the residual image to the addition unit 3 7 8. In addition, the processing of the inverse conversion unit and the inverse conversion unit 324 of the moving picture encoding device 310 ○ The image prediction unit 376 generates a microblock to be processed and outputs the predicted image to the addition unit 3 7 8. Portrait prediction i mode switching section 3 82, first portrait prediction section (first portrait 3 8 4 and second portrait prediction section (second portrait prediction means) switching section 382 is the first portrait based on the encoding mode from the decoding section 3 72 The prediction section 3 8 4 or the second image prediction section 3 8 6. The measurement section 3 84 generates a prediction image for the processing by the first image prediction processing; the second image prediction section 3 86 generates the processing envelope by the second image. Predicted portraits of image microblocks. The processing performed by these elements included in 3 76 is the same as the animation encoding ^ code section (decoded image prediction section image memory section) (the bit stream. Situation, decoding, encoding mode • :; **, and in the process of processing the target object, this reply predicts the same processing of the predicted image as 374, and the department has the prediction method) 3 86. Mode information, start the microblock image prediction processing image of the first image pre-image Prediction section device 3 1 0-55- 200529673 (52) Processing performed by the corresponding elements of the portrait prediction section 3 1 6. The addition section 3 7 8 adds the predicted portrait from the portrait prediction section 3 76 and the inverse conversion section 3 Residual Predictions of 74 Image, to generate a predicted image of the microblock of the processing object, and store it in the image memory section 3 80. Furthermore, when the specific range of the image element 値 of the image data is set, the addition unit 378 may limit the pixel 値 to Capture processing is performed within this range. The image memory section 380 stores the playback frame and the reference frame of the decoded object frame that has been generated to play the image. The reference frame is shown above, which is related to the decoding object. The playback frame of the frame with a different frame is, in this embodiment, the playback frame before one of the frames of the decoding target frame. The operation of the animation decoding device 370 will be described below. FIG. 25 shows the fourth A flowchart of a method for implementing a type-dependent animation decoding method. As shown in FIG. 25, in the decoding process of the animation decoding device 370, first, the decoding section 3 72 decodes the encoding mode information of all the microblocks of the object frame, from The bit stream is decoded (step S331). Next, the mode switching unit 382 selects a microblock as a processing object according to a specific scanning order (such as a light spot scanning order), and tests the microblock of the processing object. Whether the encoding mode is the P mode (the first portrait prediction process) (step S 3 3 2). If the encoding mode is not the P mode (No), the process moves to step S 3 38. On the other hand, the encoding mode is the P mode In case (Yes), the decoding unit 3 72 decodes the conversion information and the motion vector of the processing target microblock from the bit stream average information amount (step S3 3 3). Furthermore, the mode switching unit 3 82 starts the first The image prediction unit 384 generates a predicted image of the microblock to be processed by the first image prediction process (-56- 200529673 (53) step S 3 4). Next, the inverse conversion unit 3 74 performs inverse conversion on the conversion information to generate a revertive prediction residual image (step S3 3 5). Next, the addition unit 3 7 8 adds this to the predicted residual image and the predicted image, and generates a playback image (step S 3 3 6). This playback picture is grouped into a playback frame and memorized in the picture memory section 3 80 (step S3 3 7). Next, it is tested whether the processing of all the microblocks is completed (step S3 38). If all block processing is not completed (No), select the unprocessed microblock. And repeat the processing from step S3 32 to step S 3 3 8. • On the other hand, if all block processing ends (Yes), the process moves to step S3 39. In step S3 39, the mode switching unit 82 selects a microblock as a processing object according to a specific scanning sequence (such as a light spot scanning sequence), and tests whether the coding mode of the processing microblock is C mode. (Second image prediction processing) (Step S3 3 9). When the encoding mode is not C mode (No), the process moves to step S3 45. On the other hand, when the encoding mode is the C mode (Yes), the decoding unit 3 72 decodes the conversion information of the microblock to be processed from the bit stream average information amount (step S340). Furthermore, the mode switching unit 3 82 activates the second image prediction unit 3 86 and generates a predicted image of the processing target microblock by the second image prediction processing (step S341). Next, the inverse conversion section 374 performs inverse conversion on the conversion information, and generates a return prediction residual image (step S342). Furthermore, the addition unit 3 78 adds this to the predicted residual image and the predicted image to generate a playback image (step S 3 43). This playback picture is grouped into the playback picture frame, and is stored in -57- 200529673 (54) in the portrait memory section 380 (step S344). Furthermore, it is tested whether the processing of all the microblocks is completed (step S3 45). In the case where all microblock processing is not completed (No) ', an unprocessed microblock is selected, and the processing from step S339 to step S345 is repeatedly performed. On the other hand, if all microblock processing ends (Yes), the decoding process ends. The following describes a video decoding program that operates a computer as the video decoding device 3 70. Fig. 26 is a diagram showing a structure of a video decoding program related to a fourth embodiment and a recording medium. The animation decoding program 3 90 ′ shown in FIG. 26 is stored in a recording medium 100 and provided for use. The recording medium 100 is, for example, a recording medium such as a magnetic disk, a CD-ROM, a DVD, or a ROM, or a semiconductor memory. If the recording medium 100 is inserted into the reading device 112, the computer 110 (refer to FIG. 1 and FIG. 15) can access the animation decoding program 390 stored in the recording medium 100 from the reading device 112, and can use the animation The decoding program 390 operates as a video decoding device 3 70. As shown in Figure 15, the animation decoding program 3 90 can also be made. It is a computer data signal 130 which is superimposed on the carrier wave and is provided via the network. At this time, the computer 110 can store the animation encoding program 340 received by the communication device 124 into the memory 116 and execute the animation decoding program 390. As shown in FIG. 26, the animation decoding program 390 is provided with a main module 39 for overall processing, a decoding module 392, an inverse conversion module 3 94, an image prediction module 3 96, an addition module 3 98, and an image memory module. 400. The portrait prediction module 3 96 includes a mode switching module 402, a first portrait prediction module 404-58-200529673 (55), and a second portrait prediction module 406. Decoding module 3 9 2, reverse conversion module 3 9 4, portrait prediction module 3 9 6, addition module 3 9 8 and portrait memory module 4 0 0, mode switching module 4 0 2, first portrait prediction The module 404 and the second image prediction module 406 each execute a function of a computer, and the above-mentioned decoding section 372, inverse conversion section 3 74, portrait prediction section 3 7 6, addition section 3 7 8, image storage section 3 8 0, In the mode switching section 38, the first image prediction section 384, and the second image prediction section 386, the corresponding element functions are the same. As described above, the animation decoding device 370 can restore the animation based on the bit stream generated by the animation encoding device 3 10. In addition, the animation decoding device 370 can generate micro-blocks whose coding mode is C mode, that is, micro-blocks that generate predicted pictures by the second picture prediction process, without generating prediction assistance information such as motion vectors from the encoding side. Forecast portrait. The present invention is not limited to the above-mentioned third and fourth embodiments, and can be variously modified. For example, in the third embodiment and the second portrait prediction process, a playback frame formed by referring to a playback image of a frame to be encoded is referred to. It is also possible to replace the playback picture frame here, and the second picture prediction process can also be realized by referring to the encoding object picture frame, that is, the input picture, image body. In addition, the predicted image generated by the second image prediction processing may still be used as the playback image V. At this time, the encoded residual data of the predicted residual image formed by the difference between the predicted image and the input image generated by the second image prediction processing, It does not need to be included in the bit stream, so a bit stream with higher coding efficiency can be generated. Regarding the characteristics of animation, the part that moves more and the part that moves less -59- 200529673 (56) The background part is known. You can also save by referring to the table of the encoding mode of each microblock in advance. Determine the processing of the encoding mode. The above 'as shown in the description of the best mode of implementation of the present invention, if the present invention is used,' an image encoding device for effectively encoding an image, an image encoding method, an image encoding program, and an image encoding device which can be generated from the image encoding device of the present invention can be provided. An image decoding device, an image decoding method, and a day image decoding program for the bit stream return image. [Brief Description of the Drawings] Fig. 1 is a structural diagram showing an image encoding device related to the first embodiment. FIG. 2 is a diagram illustrating the content of a portrait prediction process. FIG. 3 is a diagram illustrating the content of a portrait prediction process. FIG. 4 is a diagram showing a description of a replacement pixel. FIG. 5 is a diagram showing a description of a replacement pixel. FIG. 6 is a diagram illustrating the content of the image correction processing. ® 7 is a flowchart showing the operation of the image encoding device for image encoding processing. _ 8 is a diagram of the structure of a recording medium representing a recording image coding program. Fig. 9 is a block diagram showing a picture decoding apparatus according to the second embodiment. •.  FIG. 10 is a flowchart showing a general operation of the image decoding process. FIG. 11 is a flowchart showing the operation of the prediction picture decoding process. FIG. 12 is a flow chart showing the operation of the pseudo-image decoding process. -60- 200529673 (57) Fig. 13 shows the structure of a recording medium for the recording image decoding process. Figure 14 is a diagram showing the hardware structure of a computer used to execute programs stored in a recording medium. Fig. 15 is a perspective view showing a computer for executing a program stored in a recording medium. . Fig. 16 is a block diagram showing a moving picture coding apparatus according to a third embodiment. FIG. 17 is a diagram illustrating processing by a coding mode determination unit. FIG. 18 is a block diagram showing the fabrication of the image prediction section. FIG. 19 is a diagram illustrating the second image prediction image processing. Fig. 20 is a flowchart showing a type-dependent animation coding method according to the third embodiment. Fig. 21 is a flowchart showing a coding mode determination process of a moving picture coding method according to the third embodiment. FIG. 22 is a flowchart showing a second image prediction process. FIG. 23 is a diagram showing the structure of the animation coding program related to the third Binsch type and the recording medium. Fig. 24 is a block diagram showing an animation decoding apparatus according to the fourth embodiment. Fig. 25 is a flowchart showing a mode-dependent animation decoding method according to the fourth embodiment. Fig. 26 is a diagram showing a common presentation of an animation decoding program and a recording medium related to the fourth embodiment. -61-200529673 (58) [Description of main component symbols] 1 Image encoding device 11 Image segmentation unit 1 2 Encoding mode determination unit 13 Predicted image generation unit 14 Subtraction unit 15 Conversion unit 16 Encoding unit 17 Inverse conversion unit 18 Addition unit 19 Memory Part T Template P Processing target pixel Q Playback pixel S Related field Y Processing target block R Object field S 1 Segmented input image S2 Determine encoding mode S3 Predictive encoding processing mode S4 Generate predicted image S5 Generate predicted residual image S6 Calculate conversion coefficient S7 Average Message Encoding-62- 200529673 (59) S8 Generate playback difference image S9 Generate playback image S 1 0 Store playback image S 1 1 Complete all blocks 101 Program area 102a Image division module 102b Encoding mode determination module 102c Prediction image generation module 102d Subtraction module 1 02e Conversion Group 1 02f encoding module 102g inverse conversion module 1 02h addition module 1 02i memory module 2 image decoding device 2 1 decoding section 22 encoding mode judgment section 23 predictive image generation section 24 reverse conversion section 25 addition section 26 memory Section 27 toggle switch 28 Complementary image generation section S20 Average information decoding-63- 200529673 (60) S30 Predictive image encoding processing S40 Complementary image encoding processing S3 1 Predictive encoding processing mode S32 Generate predictive portrait S33 Generate predictive residual image S34 Generate Play portrait S35 Store play portrait S36 Whether all blocks are finished S41 Subsequent coding processing mode S42 Generate subsequent portrait S43 Store subsequent portrait S44 Whether all blocks end 201 Program area 202 Portrait decoding program 202a Decoding model 202b Encoding mode determination module 202c Prediction image generation module 202d Conversion module group 202e Addition module group 202f. Memory module 202g switch module 202h 塡 supplemental image generation module 110 computer 100 recording media-64-200529673 (61) 1 12 reading device 114 working memory 116 memory 126 CPU 118 display 120 mouse 122 keyboard 124 Communication device 130 Computer data 0 & 3 10 Moving picture encoding device 3 12 Field segmentation unit 3 14 Encoding mode determination unit 3 16 Image prediction unit 3 18 Subtraction unit 320 Conversion unit 322 Encoding unit 324 Inverse conversion unit 326 Addition unit 328 Portrait memory section 330 Mode switching section 332 First portrait prediction section 334th '2 Portrait prediction section FI Play frame F2 Reference frame -65- 200529673 (62) S (Sc) 5301 5302 5303 5304 5305 5306 5307 5308 5309 S3 1 0 S3 1 1 S3 1 2 S3 1 3 S3 14 S3 1 5 S3 1 6 S3 1 7 S3 1 8 S3 1 9 5320 5321 START S302-1 Copy reference field segmentation field decision encoding mode encoding encoding mode information picture 1 Like predictive processing? The first image prediction processing produces a predicted residual image conversion average message encoding inverse conversion addition memory playback image Is all blocks finished? The second image prediction process? The second image prediction processing produces a predicted residual image conversion. Average message encoding Inverse conversion Addition. Playback of the image recorded in billions. Are all blocks completed? The encoding mode determination process performs the first portrait prediction process in the scan order-66-200529673 (63) S302-2 Selects the last block S3 02-3 The second portrait prediction process S302-4 Cost calculation S302-5 Selects the encoding mode S302-6 Do all blocks end? S3 02-7 Select the next block in the reverse scanning order. START 2nd image prediction processing S314-1 Select the block where no playback image is generated S3 14-2 Set the reference area S314-3 Select the unprocessed pixels in the block S3 14 -4 Setting panel · S3 14-5 Calculation related S3 14-6 Select the copy area S3 14-7 Copy the corresponding pixel in the template's unprocessed pixel copy area S314-8 Are there unprocessed pixels in the block? S314-9 Select unprocessed pixels 340 moving image encoding program 341 main module 342 domain segmentation module 344 encoding mode determination module 346 portrait prediction module 3 60 mode switching module 3 62 first portrait prediction module -67- 200529673 (64) 364 No. 348 minus 350 to 352 series 354 anti-356 plus 358 painting 370 action 372 solution 374 anti-376 painting 378 plus 380 painting 382 die S33 1 solution S332 No. S333 flat S334 No. S335 anti-S3 3 6 plus S337 save S338 S339 No. S340 Flat 2 Image Prediction Module Method Module Change Module Code Module Change Module Method Module Image Memory Module Image Decoding Device Code Section Conversion Section Image Prediction Section Law Section Image Memory Section Switching Section Code Coding Model 1 portrait prediction processing? Decoding of average message amount 1 Image prediction processing Conversion method Store and play image Are there any blocks? 2 Image prediction processing? Average Message Decoding -68-200529673 (65) S34 1 No. S3 42 No. S3 43 Plus S344 No. S345 No. 384 No. 386 No. 390 No. 39 1 Main 392 Solution No. 394 No. 396 No. 398 Picture No. 398 No. 400 No. 402 No. 404 No. 406 No. 2 Portrait prediction processing conversion method. Does the playback frame have blocks? 1Image prediction section 2Image prediction section Image decoding program module Code module-conversion module Image prediction module Method module Image memory module Switching module 1Image prediction module 2Image prediction module -69-

Claims (1)

200529673 (1) X. Patent application scope 1. An image encoding device comprising: a plurality of partial fields formed by dividing an input image of an object to be encoded into a plurality of specific sizes, and providing prediction assistance by generating a predicted image Either the first picture prediction process or the second picture prediction process required for information, generates a coding mode for predicting a picture, generates a decision means for limiting coding mode information of the coding mode, and executes a part of the aforementioned plural In the field, from the foregoing coding mode information, a part of a field in which a predicted image should be generated by the first image prediction process is limited, and a playback image that has been generated from another part of the field is extracted to generate prediction assistance information for the predicted image in that field. And generating the first portrait prediction method of the aforementioned first portrait prediction process of the predicted portrait based on the prediction assistance information, and a memorization method of memorizing the played portrait based on the foregoing portrait, and generating the encoding mode information including the previously encoded encoding information and the foregoing prediction assistance Coding means for bit stream of information. 2. The portrait encoding device described in item 1 of the scope of the patent application, wherein the aforementioned second portrait prediction processing is a template in which a part of pixels in which no prediction signal is generated is set as a template, and the previously played portrait is Reference field, and select the reference field. The field with the highest correlation with the template is the copied reference field. Pixels of the predicted signal are not generated in the template, and pixels of the corresponding pixels in the copied reference field are given. This results in the aforementioned prediction image processing. 3. The image coding-70-200529673 (2) device described in item 1 or 2 of the scope of the patent application, wherein the aforementioned determination means passes the first part of the plurality of areas to play the image through the aforementioned first image prediction processing. After generating according to a specific scanning order, select a part of the processing target area in accordance with the reverse order of the specific scanning order. Compared to the part of the processing target, the image of the front part of the scanning order is selected Compared to the partial area of the processing object, in the partial area located rearward in the aforementioned cat scan sequence, the playback portrait design of the partial area is determined by the determination of the coding area of the predicted image that should be generated by the first image prediction processing. For the aforementioned reference area, a playback image of a partial area of the processing target is generated through the second image prediction processing, and a playback portrait of a partial area of the processing target generated by the second image prediction processing is compared with the first image prediction processing. The playback image of the part of the generated processing object determines the part of the processing object. The coding mode. 4. The image encoding device described in item 1 or 2 of the scope of the patent application, wherein, in some areas further including the foregoing plural number, from the foregoing encoding mode information, it is limited that the predicted image should be generated by the second image prediction processing. The predicted image in some areas is a second image prediction means generated by the second image prediction process; the second image prediction means sets the predicted image generated by the second image prediction process as the aforementioned Play portrait. 5. The portrait encoding device described in item 1 or 2 of the scope of patent application, wherein the input portrait of the encoding object is an animation frame; in the aforementioned second portrait prediction process, the portrait of the encoding object frame is played, And at least any one of the frame-playing image-71-200529673 (3) which is given priority over the frame to be encoded, is set as the aforementioned reference area. 6. The portrait encoding device described in item 1 or 2 of the scope of the patent application, wherein the first portrait prediction processing is generated by using the aforementioned playback image in the same space as the part of the processing target to generate the prediction. Prediction image processing; In the first image prediction processing, the connection part area connecting the part of the processing target area from the coding mode is limited to the case where a part of the predicted image should be generated by the second image prediction processing. Based on the playback image of a non-connected partial field that is not connected to the partial field of the processing object, a predicted day image of the partial field of the processing object is generated. 7. The portrait encoding device described in item 6 of the scope of patent application, wherein in the first portrait prediction processing, a connection partial area that connects a partial area of the processing target from the encoding mode is limited to the second portrait Prediction processing should produce the situation of some areas of the predicted image, which are located on a straight line in the prediction direction and in the aforementioned playback portrait of the aforementioned non-connected part area in the direction of the prediction start end, based on the picture of the part of the area closest to the processing object Elementary, the aforementioned predicted picture is generated. 8. The image coding device described in item 1 or 2 of the scope of patent application, which further includes the aforementioned item by executing the aforementioned item! The predicted residual image generated by the image prediction method is calculated by calculating the difference between the predicted day image and the input image of the encoding object to generate the predicted residual image. The aforementioned encoding method encodes data based on the signal generated by the predicted residual image. , Contained in the bitstream. 9. An image encoding method, the determining means includes: for a plurality of partial fields formed by dividing an input image of an image into a specific size by dividing an encoding image of a pair of images into a specific size, determining a relevant prediction image that must be generated with prediction assistance information Which of the first image prediction process or the second image prediction process generates a coding mode for predicting a picture, and generates a decision step for defining coding mode information for the coding mode; and a first picture prediction means, which executes the aforementioned plural part In the Fen field, from the aforementioned coding mode information, < limit the part of the region where the predicted image should be generated by the first image prediction process, and extract the already-played image from the other part of the field to generate the part i field. Prediction: the prediction assistance information of the portrait, and the first portrait prediction step of the aforementioned first portrait prediction process of generating the predicted portrait based on the prediction assistance information, and the memorizing step of playing the portrait based on the image described above, and the encoding method generates A bitstream containing the data of the aforementioned encoding mode information and the aforementioned prediction subsidy information Encoding step. 1 〇 · —A kind of portrait coding program, which is functionalized by a computer. At first, for each of the plural fields formed by dividing the input image of the encoding object into a specific size, determine whether the prediction image must be generated by using prediction assistance information. Which of the first picture, the prediction process, or the second picture prediction process generates a coding mode for predicting a picture, and generates a determination means for defining coding mode information for the coding mode, and performs a part of the plural from the foregoing The coding mode information is limited to a part of a field where a predicted image should be generated by the first image prediction process, and a playback image that has been generated from another part of the field is extracted to generate prediction assistance information for the predicted image in that part of the field, and based on the prediction assistance -73-200529673 (5) Information generating the first image prediction means of the aforementioned i-th image prediction process of the predicted image, and a memory means of memorizing the played image based on the aforementioned image, and generating the information including the previously encoded encoding mode information and the aforementioned prediction. Coding method of bit stream of subsidy information. 11. An image decoding device comprising: a first image prediction process or a second image process for generating a predicted image, including a first image prediction process and a second image, which are used to generate a predicted image from a coded image divided into a plurality of partial areas formed by a specific size; Decoded encoding mode information and a bit stream of prediction assistance information that generates a predicted image based on the first portrait prediction process, and decodes the encoding mode information and the prediction assistance information in a decoding method; and the aforementioned partial domain / domain, From the foregoing / mentioned decoding mode information, the prediction picture of a part of the area in which the predicted picture should be generated by the first picture prediction process is limited, and the generated picture is played back, and the first picture i prediction process using the aforementioned prediction assistance information is limited. The first image prediction means and the plurality of partial fields described above limit the predicted image in a part of the field where the predicted image should be generated by the second image prediction processing from the aforementioned decoding mode information, and by the second image prediction The second image prediction means generated by the image prediction processing, and the playback image based on the predicted image is stored. Memory means; in the aforementioned second image prediction processing, a field in which pixels that do not generate a prediction signal are a part is set as a template, and the playback image stored in the memory means is set as a reference field, and the reference field and the foregoing are selected Template-74- 200529673 (6) The most relevant field is the copy reference field. On pixels that do not generate the aforementioned prediction signal in the aforementioned template, the pixels of the corresponding pixels in the aforementioned duplicated reference field are given, thereby The aforementioned predicted portrait is generated. 1 2 · The image decoding device described in item 11 of the scope of the patent application, wherein the aforementioned determination means i is used to limit the portion of the image that can be predicted by the first image prediction processing from the aforementioned solution mode information. Field, in accordance with a specific scanning sequence, the first predicted image is generated by the first image predicting location, and the playback image generated based on the predicted image is stored in the memory means, and the second image is interpreted by the military measurement means. The decoded presentation information is limited to be produced by the second image prediction process. Part of the field of the predicted image field, the aforementioned predicted image is generated by the second image prediction process in accordance with the aforementioned specific scanning order. 1 3 · The portrait decoding device described in item i 1: of the scope of patent application; •. '· ....-' * Among them, the second portrait prediction means will generate the aforementioned predicted image by the aforementioned second image prediction process. , Set to the aforementioned play image. 1 4. The portrait decoding device described in the application claim # 1, wherein the portrait of the decoding object is an animation frame;: 1, '; H.L The foregoing second image prediction means is stored in the foregoing At least one of the playback image of the aforementioned solution object frame and the playback image of the frame processed preferentially over the decoding object frame is set as the aforementioned reference area. _ 1 5 · The portrait decoding device described in item 11 of the scope of patent application, in which the first and first image prediction processing are generated by using the aforementioned playback image in prediction in the same space as the part of the aforementioned processing object. Pre-test-75- 200529673 (7) Image processing; In this first image prediction processing, from the aforementioned encoding mode, the connected partial areas connected to the aforementioned areas of the physical object are limited to the previous second image prediction processing In the case where a part of the predicted image is to be generated, a portrait is taken in a non-connected part of the field that is not connected to the part of the processed object, and a predicted image of the part of the processed object is generated. 16. The image decoding device described in item 15 of the scope of the patent application, wherein in the first image prediction processing, the predicted information is limited from the decoding-type information and the second image prediction processing should generate a predicted image. In the field, in the aforementioned play portrait in the aforementioned non-connected part field located in the prediction direction and at the beginning end of the pre-Jew, the predicted residual image is generated based on the pixel 値 of the pixel in the part field closest to the processing object. 1 7. The portrait decoding device described in item 11 of the scope of the patent application, wherein the bit stream includes a signal of a predicted residual picture generated by a calculation based on a difference between the pre-image in the foregoing partial field and the portrait in the partial field. The generated data; the aforementioned decoding means is included in the data of the aforementioned bit stream, 'the data formed by coding the signal generated based on the predicted residual image, and the signal is coded; the image decoding device is further provided with an addition based on The playback predicted image generated by the signal of the decoding means code and the predicted predicted image are generated, thereby generating a playback image generating method of the playback image. 18. —A kind of portrait decoding method 'equipped with: Describes the resolution of the broadcast sub-division retention image code-76- 200529673 (8) For each partial area of the complex number formed by dividing the decoded image into a specific size' Decoding means including a bit stream for decoding first picture prediction processing or second picture prediction processing used to generate a predicted picture, and a bitstream including decoding assistance information for generating a predicted picture by the first picture prediction process. Decoding step of decoding the encoding mode information and the prediction assistance information, and the first picture prediction means, among the plurality of areas of the plurality, from the foregoing decoding mode information, the area in which a pre-beach picture should be generated by the first picture prediction processing is limited. The prediction image of a part of the field is generated by the first image prediction step and the second image prediction method by the first image Jia Xuli 'using the aforementioned prediction assistance information. The aforementioned decoding mode information is limited to the pre-I of a part of an area where a predicted image should be generated by the aforementioned second image prediction process. Image, the 2nd image ^ prediction step generated by the 2nd image prediction process, ... and the method of memorizing the memory step of playing the image based on the predicted image; in the 2nd image prediction process, no prediction signal is generated. The area where a pixel is a part is set as a template, and the aforementioned playback image stored in the aforementioned memory means is set as a reference area, and the area with the highest correlation with the template in the reference area is selected as the copy reference area, and the aforementioned area is not generated in the aforementioned template. On the pixels of the prediction signal, the pixels of the corresponding pixels in the copied reference area are applied to generate the aforementioned predicted portrait. 19. An image decoding program, which uses computer functions as: -77- 200529673 (9) For each part of a complex number formed by dividing an image to be decoded into a specific size, it is adopted from the limitation of encoding to the first to generate a predicted image. 1 picture prediction processing 'or 2nd picture prediction processing decoding mode information, and a bit stream of prediction assistance information for predictive pictures generated by the first picture prediction processing, and decoding means decodes the encoding mode information and the prediction assistance information. From the aforementioned decoding mode information, the "decoding means" and the above-mentioned plural partial fields "limit the predicted images in some areas of the fields where the predicted images should be generated by the first image prediction process, and the Jti described the first section of the prediction assistance information. The first image prediction means generated by the 1 image prediction process and the plurality of partial fields described above limit the predicted images in a part of the fields in which the predicted image should be generated by the second image prediction process from the aforementioned decoding mode information. The second image prediction means generated by the first image prediction process, and the memory is based on the previous In the second image prediction processing, in the aforementioned second image prediction processing, a field in which pixels that do not generate a prediction signal are a part is set as a template, and the playback image stored in the storage means is set as a reference field, and is selected. The field with the highest correlation with the template in the reference field is the copy reference field. On pixels that do not generate the prediction signal in the template, pixels of the corresponding pixels in the copy reference field are applied to generate the aforementioned pixels. Predict portraits and make computers functional -78-